Geochemical Studies of Tektites from East Asia

Home , Australasian strewnfield, Tektite, Zhamanshin crater

Geochemical Journal, Vol. 38, pp. 1 to 17, 2004

Geochemical studies of tektites from East Asia

YUNG-TAN LEE,1 JU-CHIN CHEN,1* KUNG-SUAN HO2 and WEN-SHING JUANG2

1Institute of Oceanography, National Taiwan University, Taipei, Taiwan 10617, R.O.C. 2Department of Geology, National Museum of Natural Science, Taichung, Taiwan 40419, R.O.C.

(Received March 3, 2003; Accepted June 18, 2003)

Thirty tektites from East Asia (including Wenchang and Penglei of Hainan Island, Maoming of Guandong, China; Khon-Kaen of Thailand; Bao Loc of Vietnam; Rizal of Luzon, Philippines) have been analyzed for major and trace element contents and Rb-Sr isotopic compositions. All the samples studied are splash form tektites. The trace element ratios Ba/Rb (avg. 3.74), Th/Sm (avg. 2.31), Sm/Sc (avg. 0.43), Th/Sc (avg. 0.99) and the rare earth elements (REE) contents of tektites studied are similar to the average upper continental crust. The chemical data of tektites in this study indicate that they were derived from similar target rocks which may be related to post-Archean upper crustal materials. The tektites from East Asia have high positive εSr(0) values-ranging from 164.2 to 198.6, indicating that they were not dominantly derived from recent young sediments, such as soil or loess. The Ar-Ar ages (736.8 ± 55.5~814.6 ± 24.4 ka) of tektites of the present study are consistent with the age of other Australasian tektites, which indicates that all Australasian tektites were derived from a single impact event. Previous studies (Koeberl, 1992; Blum et al., 1992; Schnetzler, 1992) and the present chemical data suggest that these tektites are the result of melting at a single site, which is most probably located in the southern part of the Thailand-Laos border. Mixing calculations based on the model suggested by Ho and Chen (1996) for various amounts and combinations of target rocks indicate that the best fit for East Asia tektites is a mixture of 61% greywacke, 32% sandstone and 7% shale.

Keywords: tektites, geochemistry, strewn field, East Asia, 40Ar-39Ar dating

Apart from the microtektites found in deep-sea INTRODUCTION sediments (Glass, 1990; Lee and Wei, 2000), tektites on Tektites are relatively homogeneous, dense glass ob- land can be subdivided into three groups: (a) normal or jects found scattered over large areas of the Earth’s sur- splash form tektites, (b) aerodynamically ablated tektites, face called strewn fields. Four strewn fields are known and (c) Muong Nong-type tektites (layered tektites) common names and approximate ages are: North (Koeberl, 1992). The first two groups are only slight dif- American, 34.9 Ma (Storzer and Wagner, 1971); Central ferent in their appearance and physical characteristics. European, 14.5 Ma (Schwarz and Lippolt, 2002); Ivory The shapes of splash form tektites (spheres, droplets, tear- Coast, 1.07 Ma (Koeberl et al., 1997) and Australasian, drops, dumbbells etc.) are often erroneously described as 0.77 Ma (Izett and Obradovich, 1992). Tektites found aerodynamical ablated tektites (flanged button). Muong within a given strewn field are related with respect to their Nong-type tektites are named after a locality in Laos physical and chemical properties besides their age. The where they were first found by Lacroix (1935). They are smallest strewn field is the Central European, and the larg- larger in size, less homogeneous, having higher abun- est is the Australasian field. Some of the strewn fields dances of volatile elements (e.g., B, Cu, Zn, Ga, As, Se, can be divided into different substrewn fields. For exam- Sb and Pb) and water and contain more bubbles and some ple, some of the North American tektites can be relict minerals (e.g., coesite, zircon, corundum, rutile, subclassified as bediasites (Texas) and georgiaites chromite, etc.) which indicate a sedimentary rock as the (Georgia). The Central European strewn field includes source rock (Koeberl, 1992). Besides, based on their tektites named moldavites from Moravia and Bohemia, heavier weights and irregular shapes, some authors whereas the Australasian strewn field consists of several (Stauffer, 1978; Hartung and Rivolo, 1979) suggested that subfields. Muong Nong-type tektites probably have not traveled far from their location of origin, and may therefore be close to the impact crater. The origin of tektites was a scientific puzzle for many *Corresponding author (e-mail: [email protected]) years. Although O’Keefe (1976, 1994) favored a lunar Copyright © 2004 by The Geochemical Society of Japan. volcanic origin, most authors accept that tektites are the

1 product of hypervelocity impact on Earth and represent The purpose of the present study is to analyze the melted target rock ejected during crater formation (Glass, major and trace elements (including rare-earth elements, 1990; Wasson, 1991; Blum et al., 1992; Koeberl, 1994). REE) and the Rb-Sr isotopic ratios and to date the tektites Recently, Futrell (2000) made a case for lunar provenance, by Ar-Ar method in order to shed some light on the ori- based on the theory that tektites could have erupted as gin of the tektites, their parent material and the possible volcanics at the lunar surface, but the case has been re- source crater. torted by Glass (2000), who argued that the silica-rich igneous material in lunar samples is miniature and that it SAMPLES AND ANALYTICAL METHODS is petrographically and compositionally distinct from tektite glass. The tektites were sampled personally by K. S. Ho and Lead isotope data (Wampler et al., 1969) indicate that W. S. Juang from Wenchang (5 samples) and Pengli (5 the parent material of tektites is terrestrial, and consist- samples) of Hainan Island, Maoming (5 samples) in ent with young sediments, including oceanic sediments. Guandong, China, Khon-Kaen (5 samples) in Thailand, Based on 10Be contents of tektites, several authors (Blum Bao Loc (5 samples) in Vietnam and Rizal (5 samples) in et al., 1992; Koeberl, 1992) favor a sedimentary origin Luzon, Philippines (Fig. 1). All the tektite samples are for the parental material of Australasian tektites. Accord- complete pieces and their weights ranging from 3.88 g to ing to 10Be contents (about 1 × 108 atoms) of 7.09 g for Wenchang, 1.64 g to 9.27 g for Pengli, 34.8 g Australasian tektites, Pal et al. (1982) suggested that cos- to 74.5 g for Maoming, 11.45 g to 17.83 g for Khon-Kaen, mic-ray bombardment of the australites cannot produce 20.64 g to 30.02 g for Bao Loc and 11.4 g to 59.3 g for the measured amounts of 10Be either at the earth’s sur- Rizal. The 30 tektites samples in the present study all face or in space, indicating a sedimentary precursor that belong to the splash-form and they are oval, elongated or adsorbed from precipitation 10Be produced in the atmos- dumbbell-shaped. phere. Thirty samples were cleaned with distilled water and

Fig. 1. Localities of analyzed East Asia tektites (stars) and the distribution of Jurassic exposures in Southeast Asia: 1, hachured areas denote marine Jurassic exposures; 2, stippled areas show nonmarine Jurassic exposures (after Sato, 1992).

2Y.-T. Lee et al. acetone in ultrasonic cleaner and crushed into chips by a total of the peak heights and their errors from the square hammer while the samples were wrapped with plastic root of the sum of squares of the peak height errors, for sheets. Several larger glass chips were selected and ground all of the temperature steps. The detailed results of 40Ar/ into powders in an agate mortar. 39Ar laser fusion experiments are listed in Table 5. The chemical analyses of tektites from East Asia have been carried out by colorimetry (Si, Al, Ti, P), atomic CHEMICAL CHARACTERISTICS OF absorption (Fe, Mg, Ca, Na, K, Mn) and inductively cou- THE TEKTITES FROM EAST ASIA pled plasma mass spectrometry (Ba, Co, Cr, Cs, Cu, Hf, Ga, Li, Nb, Ni, Rb, Sc, Sr, Ta, Th, U, V, W, Y, Zr, Zn and Major element composition REEs) at the National Taiwan and Tsing-Hua Universi- Koeberl (1990) indicated that tektites are character- ties. ized by high SiO2 content (65 to 85 wt%) and rather high Calibration curves were constructed using U.S.G.S. FeO and low alkali contents. Most tektites from East Asia standard rocks AGV-1, BCR-1, W-2, and G2 and NBS have SiO2 contents ranging from 71 to 81 wt% which is 278 standard obsidian. Values for these rock standards consistent with previous observation on Australasian were adapted from compilations by Govindaraju (1994). tektites (Table 1). Based on the high abundance of silica The precision of the analyses in the present study are as in the tektites and the occurrence of 10Be in the tektites follows: ΣFeO ± 1.46%, MgO ± 1.14%, CaO ± 1.79%, (Tera et al., 1983; Englert et al., 1984; Pal et al., 1982), Na2O ± 0.95%, K2O ± 1.58%, MnO ± 1.88%, Cr ± 1.93%, the lunar volcanic origin for tektites may be excluded. Cu ± 1.04%, Zn ± 1.37%, Rb ± 1.95%, Sr ± 1.16%, The major oxide composition of tektites from different Y ± 1.99%, Nb ± 4.03%, Ba ± 1.14%, La ± 1.31%, localities in the present study, except Al2O3 and CaO, are Ce ± 1.47%, Nd ± 2.65%, Sm ± 1.45%, Eu ± 2.15%, generally uniform. The relative depletion of Al2O3 in Tb ± 3.19%, Yb ± 2.06%, Lu ± 3.45%, Hf ± 4.18% and tektites from Maoming, Guandong, China (Al2O3 aver- Th ± 3.02%. aging 7.06 wt%) and Rizal, Luzon Philippines (Al2O3 Seven tektites (TK-2, W-1, W-2, W-3, BL-12, BL-14 averaging 6.07 wt%) may be related to the lower content and BL-20) were selected for Rb-Sr isotopic composi- of shale in the source rocks. In addition, the variable con- tion analyses. Each tektite (about 2 gm) after cleaning tents of CaO in the East Asian tektites may be due to dif- with acetone and distilled water, was individually dis- ferent amounts of limestone involved during the forma- 84 87 solved in a HF-HNO3 mixture; and tracers ( Sr and Rb) tion of the tektites. addition and chemical separations followed. Isotopic com- Except for higher Na2O, the average chemical compositions of the Rb and Sr were measured using a Finnigan positions of tektites in this study closely resemble that of MAT 261 mass spectrometer at AMDEL Company, splash-form indochinites (Table 2). Except for Na2O, CaO Thebarton, Australia. and K2O the concentrations of the major oxides are simi- Ages of five representative tektites (W-1, W-2, KK-1, lar to those of the upper continental crust (Fig. 2) (Taylor M-01 and M-03) were analyzed by the Ar-Ar method. and McLennan, 1985). The relatively lower Na2O, CaO About 1.5 to 3 gm of 40Ð120 mesh from each sample was and K2O contents found in the tektites of this study sug- wrapped in aluminum foil packets and stacked in gest that the source material probably contains relatively aluminum canister with LP-6 biotite standard (127.7 ± low contents of carbonate and feldspar. Plots of total FeO 1.4 Ma, Odin, 1982) and HDB-1 biotite (24.7 ± 0.4 Ma, versus other major oxides allow the tektites from differ- Fuhrmann et al., 1987) and irradiated in the Open Pool ent substrewn fields to be distinguished (Koeberl, 1990) Reactor at National Tsing-Hua University with neutron (Fig. 3). flux of 1.566 × 1013 n/cm2 sec for 8 hours. After irradiation, samples were loaded in degassed fused silica boats, Trace element composition and placed into a degassed fused quartz tube, baked at Except for Co, Ga and Sc, the average trace element 250°C for 24 hours and then incrementally heated fol- contents of the tektites in this study closely resemble the lowing a 30-min. per step schedule using a resistance fur- average composition of previously analyzed splash-form nace. The purified gas was analyzed with a Varian-MAT indochinites (Fig. 4). Comparison of trace element data GD 150 mass spectrometer. The concentrations of 36Ar, between tektites in this study and the average upper crust 37Ar, 38Ar, 39Ar and 40Ar were corrected for system blanks, (Fig. 5) suggests that some trace elements correlations for radioactive decay of the nucleogenic isotopes and for are not close (i.e., Cr, Li, Sc, Co, Hf and Cs). However, minor interference reactions involving Ca, K and Cl, fol- Ba, Rb, Ga and U are closely correlated between the lowing procedures outlined in detail by Lo and Lee (1994). tektites analyzed and the upper continental crust. A weighted mean of J values obtained from the analyses Most scientists favor a terrestrial sedimentary precur- of irradiation standard minerals is adopted in the age cal- sor material for tektites (Taylor, 1973; Koeberl, 1986, culations. Integrated dates were calculated from the sum 1992). Sedimentary rocks have higher Th/U ratios than

Geochemical studies of tektites 3 Asia tektites

able 1. Major and trace element compositions of 30 East

4Y.-T. Lee et al. n.d. = not determined.

Geochemical studies of tektites 5 Table 2. Major and trace element compositions of average East Asia tektites compositions compared with average Muong Nong-type and splash-form indochinites, and average upper continental crust

A BCDEFGHI J

[wt%]

SiO 2 72.49 73.06 73.25 72.93 78.53 75.87 74.36 78.93 72.70 66.00

Al2O3 13.73 13.54 13.75 12.61 7.06 6.07 11.13 10.18 13.37 15.20 MgO 3.38 1.70 2.69 2.62 2.31 2.87 2.59 1.43 2.14 2.20 ΣFeO 5.16 4.55 5.15 4.52 5.11 5.45 4.99 3.74 4.85 4.50 CaO 1.10 1.97 0.43 1.76 3.01 4.58 2.14 1.21 1.98 4.20

Na2O 1.40 1.53 1.40 1.32 1.30 1.62 1.43 0.92 1.05 3.90

K2O 2.30 2.61 2.15 2.45 2.27 2.54 2.39 2.42 2.62 3.40 MnO 0.10 0.09 0.10 0.11 0.11 0.11 0.10 0.08 0.08 0.08

TiO2 0.75 0.77 0.74 0.83 0.78 0.82 0.78 0.63 0.78 0.50

P2O5 0.09 0.09 0.04 0.04 n.d. n.d. 0.04 n.d. n.d. n.d. Total 100.49 99.98 99.71 99.20 100.49 99.92 99.97 99.54 99.57 99.98 [ppm] Li 49.0 55.6 54.1 42.7 45.1 61.7 51.4 42.1 47.1 20.0 Sc 10.6 13.7 13.4 9.9 17.8 39.4 17.5 7.7 10.5 11.0 V8199806695668181 72 63 60 Cr 91.4 85.1 77.3 53.0 67.8 103.5 79.7 60.6 63.0 35.0 Co 13.0 15.9 11.9 10.6 23.1 30.2 17.4 12.6 11.0 10.0 Ni 29.1 36.4 20.1 n.d. n.d. n.d. 14.3 48.6 19.0 20.0 Cu 4.72 10.02 5.56 5.80 n.d. n.d. 4.4 14.3 4.0 25.0 Zn 21 36 18 10 17 26 21 67 6 71 Ga 11.9 11.8 10.1 6.7 18.0 24.0 13.7 24.2 8.2 17.0 Rb 109 133 130 97 103 118 115 110 130 112 Sr 150 147 134 141 141 183 150 135 90 350 Y 37.4 39.1 41.2 27.4 48.4 90.9 47.4 n.d. n.d. n.d. Zr 265 323 273 245 429 974 418 280 252 190 Nb 17.9 16.4 17.9 12.9 25.0 39.3 21.6 n.d. n.d. n.d. Cs 6.42 8.00 6.68 5.76 5.96 9.88 7.12 5.09 6.50 3.70 Ba 389 385 419 370 417 603 430 341 360 550 La 40.4 41.5 40.3 35.2 49.8 86.3 48.9 28.2 36.5 30.0 Ce 84.0 87.6 80.7 72.1 97.6 157.0 96.5 60.7 73.1 64.0 Nd 37.2 37.7 37.8 29.9 41.6 51.4 39.2 29.1 33.2 26.0 Sm 7.5 7.5 7.2 5.9 7.4 9.2 7.5 4.9 6.6 4.5 Eu 1.28 1.21 1.24 0.98 0.50 0.68 0.98 1.01 1.22 0.88 Gd 5.73 5.56 5.58 5.40 5.25 6.81 5.72 4.30 5.24 3.80 Tb 0.95 0.94 0.95 0.83 0.64 0.99 0.88 0.75 0.85 0.64 Dy 5.83 5.98 5.68 4.47 n.d. n.d. 3.66 4.75 5.58 3.50 Yb 2.87 3.21 2.89 2.46 2.82 3.92 3.03 2.71 2.90 2.20 Lu 0.50 0.48 0.49 0.40 0.47 0.68 0.50 0.42 n.d. 0.32 Hf 6.63 7.92 6.65 6.81 9.21 12.66 8.31 8.13 6.95 5.80 Ta 1.6 0.6 1.5 1.0 1.3 1.8 1.3 1.17 1.6 2.2 W 0.70 1.15 0.45 0.79 n.d. n.d. 0.52 1.02 0.29 2.00 Th 14.2 16.8 14.0 12.9 18.1 27.2 17.2 11.1 14.0 10.7 U 2.34 2.50 2.17 2.04 2.35 4.47 2.65 2.48 2.07 2.8

A, Average Wenchang tektites (China); B, Average Penglei tektites (China); C, Average Khon-Kaen tektites (Thailand); D, Average Bao Loc tektites (Vietnam); E, Average Maoming tektites (China); F, Average Rizal tektites (Philippines); G, Average 30 East Asia spalsh form tektites (this study); H, Average Muong Nong-type indochinites (Koeberl, 1992); I, Average splash-form indochinites (Koeberl, 1992); J, Average upper continental crust (Taylor and McLennan, 1985). n.d. = not determined.

igneous rocks (McLennan and Taylor, 1980). The high lish that the geochemistry of tektites is almost identical Th/U ratios (>6) found in tektites of this study indicate to the composition of the terrestrial upper crust. We use that sedimentary rocks may be the major source materi- the Ba/Rb, Sm/Sc, Th/Sm and Th/Sc ratios to evaluate als of these tektites. The ratios of Ba/Rb, K/U, Th/Sm, whether the tektite compositions are compatible with a Sm/Sc and Th/Sc were used by Koeberl (1994) to estab- oceanic crustal, lower continental crustal or upper conti-

6Y.-T. Lee et al. Fig. 2. Comparison of average major element compositions of East Asia splash form tektites (this study) with average compositions of upper continental crust (Taylor and McLennan, 1985).

Fig. 4. Comparison of trace element compositions of East Asia splash form tektites (this study) with (A) average trace element compositions of splash-form indochinites, (B) average trace element compositions of Muong Nong-type indochinites (Koeberl, 1992).

nental crustal source. The average element ratios of tektites of the present study (Ba/Rb = 3.74, Th/Sm = 2.31, Sm/Sc = 0.43, Th/Sc = 0.99) (Table 3) are similar to those of average upper continental crust. The lunar provenance of tektites suggested by Futrell (2000) is unlikely (Fig. 6), on the other hand, a La vs. Th plot (Fig. 7) indicates that the tektites in this study cluster around the correlation line defind by post-Archean sediments (La/Th = 2.8) and differ from the Archean sediment correlation line (La/ Th = 3.5; Taylor and McLennan, 1985) which indicate that these tektites may be derived from post-Archean sedimentary rocks such as sandstones, greywacke and shale.

Fig. 3. Plots of total Fe (as FeO) versus major elements for The rare-earth elements composition splash form tektites from East Asia (this study) and Muong The rare-earth elements (REEs) are among the most Nong-type tektites (Koeberl, 1992). important trace elements useful in petrogenetic studies of the parent rocks because their patterns are not affected by melting or metamorphic processes. Impact processes

Geochemical studies of tektites 7 Fig. 5. Comparison of average trace element composition of East Asia splash form tektites (this study) with average trace element composition of upper continental crust (Taylor and McLennan, 1985).

do not alter the REE patterns significantly, with the possible exception of affecting the most volatile REEs if frac- tional vaporization took place (Koeberl, 1990). Several authors (Taylor and McLennan, 1979; Koeberl, 1990, 1992; Wasson, 1991; Ho and Chen, 1996) concluded that the LREE in the Australasian tektites are strongly enriched relative to chondritic abundances and that the REE patterns of these tektites are similar to those of upper crustal sedimentary rocks. The REE patterns for tektites in this study are similar Fig. 6. Sm vs. Th and Rb vs. Ba plots for various tektites from to those of previously analyzed splash-form indochinites East Asia (this study). The “Earth” line has been calculated from average upper crustal material (Koeberl and Fredriksson, (Fig. 8), indicating that they are all derived from similar 1986). East Asia tektites (this study) clearly show a terrestrial parent rocks of upper continental crust affinity. The upper crustal ratio. Symbols same as in Fig. 3. present authors further excluded the lower crust and oceanic crust as source rocks of these tektites due to their different REE patterns (Fig. 9). We suggest that Rizal splash form tektites may have more Regional compositional variations contribution from heavy minerals such as zircon and The upper continental crust (UCC, Taylor and chromite. The former is enriched in Zr, Ta, Th and REE, McLennan, 1985) normalized compositional distributions while the later is enriched in Cr and Fe. This is consistent of splash form tektites from East Asia (data from Table with the relatively higher total FeO content found in Rizal 2) are shown in Fig. 10. It should be noted SiO2 and K2O splash form tektites averaging 5.45 wt% (Table 2). show relatively restricted variation indicating that the contribution of quartz and K-feldspar are relatively con- Isotopic composition stant. MgO and CaO show some variations indicating that Nd and Sr isotopic studies of Australasian tektites pro- the contribution of parental carbonate rocks may be vide information on the age and provenance of the target slightly variable. materials and allow us to characterize the target area and For the trace elements, Rizal, Philippine tektites have the impact process leading to tektite formation (Blum et higher Sc, Th, Cr, Zr, Ta, La and Yb as compared with al., 1992). Shaw and Wasserburg (1982) showed that Penglei (China), Wenchang (China), Maoming (China), tektites from each strewn field have well-defined ranges Khon-Kaen (Thailand) and Bao Loc (Vietnam) tektites. of Nd and Sr isotopic compositions. Each tektite group is

8Y.-T. Lee et al. Table 3. The average Ba/Rb, Th/Sm, Sm/Sc and Th/Sc ratios of tektites from each location and the average of thirty East Asia splash form tektites (this study) compared with average upper continental crust (UCC, Taylor and McLennan, 1985)

WenchangPenglei Khon-Kaen Bao Loc Maoming Rizal Avg. East UCC China China Thailand Vietnam China Philippines Asia tektites

Ba/Rb 3.57 2.95 3.24 3.83 4.04 5.13 3.74 4.91 Th/Sm 1.90 2.25 1.94 2.18 2.46 2.95 2.31 2.38 Sm/Sc 0.71 0.54 0.54 0.60 0.43 0.23 0.43 0.41 Th/Sc 1.34 1.23 1.05 1.31 1.06 0.69 0.99 0.97

Fig. 8. Chondrite-normalized REE diagram of East Asia splash form tektites (this study, average values) and various materi- Fig. 7. Correlation diagram of La vs. Th for East Asia tektites. als. Cl-normalized values after Evensen et al. (1978); data for The data points for East Asia splash form tektites (this study) Muong Nong-type and splash-form indochinites from Koeberl cluster around the correlation line defined for post-Archean (1992); upper crust from Taylor and McLennan (1985). sediments (data for Archean and post-Archean sediments from Taylor and McLennan, 1985). Symbols same as in Fig. 3.

strewn field are characterized by quite variable εSr(0) iso- Nd characterized by a uniform Nd model age, TCHUR , inter- topic composition (Table 4; Fig. 11). Wasson (1987, 1991) preted as the time of formation of the crustal segment suggested that the source of tektites may be a soil such as which weathered to form the parent sediment of the loess which has a homogeneous Sr isotopic composition tektites: (1) ~1.15 AE for Australasian tektites; (2) ~1.91 (Taylor et al., 1983). The variable Sr isotopic composi- AE for Ivory Coast tektites; (3) ~0.9 AE for moldavites; tions found in the Australasian tektites may be used as (4) ~0.65 AE for North American. Several authors criteria to exclude soil or loess as the dominant parental (Compston and Chapman, 1969; Taylor, 1969; Shaw and source rocks. Wasserburg, 1982) previously suggested that the age of The Sr isotopic data obtained by the present study the Australasian tektite source material is around 200 to support the conclusion reached by Blum et al. (1992) (Fig. 400 Ma. Based on the Nd and Sr isotopic data, Blum et 12), since our data all fall within the wedge-shaped array al. (1992) revealed that the Australasian tektites were defined by all Australasian tektites. derived dominantly from a sedimentary formation with a The age of the Australasian impact is known to be narrow range of stratigraphic ages, close to 170 Ma about 0.77 Ma based on radiometric ages of the tektites (Jurassic), by a single impact event. and the stratigraphic position of microtektites found in Because of the wide age range and Rb/Sr ratios of deep-sea sediments of the region (Gentner et al., 1967; crustal rocks, the continental crust has a heterogeneous Glass, 1979; Burns and Glass, 1989; Lee and Wei, 2000). Sr isotopic composition. Tektites from the Australasian The Ar-Ar ages of tektites obtained by the present study

Geochemical studies of tektites 9 range from 736.8 ± 55.5 to 814.6 ± 24.4 ka (Table 5), pact craters: the Central European with the Ries crater in which are consistent with the published ages (Burns and Germany at 14.5 Ma (Schwarz and Lippolt, 2002) and Glass, 1989; Wasson, 1991; Lee and Wei, 2000) of the the Ivory Coast strewn field with the Bosumtwi crater in Australasian tektites indicating that the Australasian Ghana at 1.07 Ma (Koeberl et al., 1997). And the North tektites are or probably are derived from a single impact American may be associated with the Chesapeake Bay event.

SOURCE REGION FOR THE AUSTRALASIAN TEKTITE STREWN FIELD The Australasian tektites strewn field is the youngest and the largest one on earth which covers more than one- tenth of the Earth’s surface (Glass and Pizzuto, 1994). Three of the four recognized strewn fields have been chemically and geochronologically associated with im-

Fig. 9. Average REE composition of East Asia splash form Fig. 10. Upper continental crust (UCC, Taylor and McLennan, tektites (this study) and various source materials normalized 1985) normalized compositional distributions of splash form to Cl-chondrite (Evensen et al., 1978). Data for upper conti- tektites from East Asia (data from Table 2). P: Penglei, China; nental crust from Taylor and McLennan (1985), lower crust W: Wenchang, China; M: Maoming, China; KK: Khon-Kaen, from Rudnick (1992), oceanic crust from Hofmann (1988). Thailand; BL: Bao Loc, Vietnam; R: Rizal, Philippines.

Table 4. Summary of Rb-Sr results and model age calculations

87 86 87 86 Sr UR Sample number Location Rb (ppm) Sr (ppm) Rb/ Sr Sr/ Sr ε fRb /Sr TSr (Ma)

W-1 Wenchang 108.1 159.8 1.960 0.716771 ± 6 174.2 22.70 460 W-2 Wenchang 109.3 159.2 1.989 0.716818 ± 6 174.8 23.05 455 W-3 Wenchang 102.4 155.9 1.903 0.717082 ± 10 178.6 22.01 487 TK-2*Penglei 123.2 131.6 2.706 0.718490 ± 36 198.6 31.72 376 BL-12Bao Loc 101.0 137.0 2.207 0.716507 ± 16 170.4 25.69 398 BL-14Bao Loc 98.0 156.0 1.793 0.716066 ± 15 164.2 20.68 476 BL-20Bao Loc 97.0 134.0 2.846 0.716974 ± 27 177.1 33.41 318

Ratios (87Sr/86Sr) normalized to 87Sr/86Sr = 8.3752 and using 85Rb/87Rb = 2.600; λ(87Rb) = 1.42 × 10Ð11 yrÐ1. Sr UR 87 86 87 86 ε , fRb/Sr and TSr are calculated relative to the present day bulk earth composition of Rb/ Sr = 0.0827, Sr/ Sr = 0.7045 and using λ(87Rb) = 1.42 × 10Ð11 yrÐ1. *From Ho and Chen (1996).

10 Y.-T. Lee et al. Fig. 11. Histogram of εSr(0) for East Asia tektites (this study and Ho and Chen, 1996) showing ranges for Sr isotopic compositions observed within the Australasian strewn field. Data on old continental crust, mid-ocean ridge basalts (MORB), continental flood basalts (CFB), oceanic crust, indochinite and australites from Shaw and Wasserburg (1982) and Blum et al. (1992).

Fig. 13. Correlation plots of compositional data from the mixing model vs. compositional data for East Asia splash form tektites. Mixture used for this model is 61% greywacke, 32% sandstone and 7% shale (see Table 6 for compositional data).

Sr Rb/Sr Fig. 12. Plot of T UR vs. 1/f for Australasian tektites (this study and literature values by Blum et al., 1992, Ho and Chen, 100 km. Based on masses of Australasian microtektites, 1996). The Y-intercept gives the time of last Rb-Sr fractionation Glass (1993, 2003) suggested that the minimum diam- and corresponds to the time of sedimentation of a sedimentary eter of source crater for Australasian tektites is around parent material (Blum et al., 1992). 40 km. Elgygytgyn Crater, Zhamanshin Crater and a structure in Cambodia with a diameter of 10 km may be excluded as source crater. Geographic variations in number, size, shape, petrography and composition of the Australa- structure (Poag et al., 1994; Koeberl et al., 1996). The sian tektites suggest that the source region may be in site of the impact for the Australasian strewn field has Indochina (Barnes, 1961; Chapman and Scheiber, 1969; not yet been determined. Blum et al., 1992; Schmidt et al., 1993; Glass and Pizzuto, Bentley (1979) concluded that there was no giant 1994). meteorite crater in Wilks Land. Elgygytgyn Crater has a Based on the fluence of microtektites (number per cm2 diameter of 18 km and an age of 3.50 ± 0.50 Ma and of the column) recovered from marine cores from oce- Zhamanshin Crater has a diameter of 13.5 km and an age anic regions near southeast Asia, the East Indies and Aus- of 0.7Ð1.0 Ma (Matsubara et al., 1991). Blum et al. (1992) tralia, Schmidt et al. (1993) concluded that microtektite and Koeberl (1994) considered that the source crater for fluences tend to decrease with increasing distance from Australasian strewn field should have a diameter of 50 to southeast Asia. Based on the microtektite concentrations

Geochemical studies of tektites 11 Table 5. The results of 40Ar/39Ar laser fusion experiments on representative tektite samples from East Asia

No. Atmos. (%) 36Ar/39Ar 37Ar/39Ar 38Ar/39Ar 40Ar/39Ar 40Ar/36Ar Date (ka)

W-1 (Wenchang, China) 405 95.76 0.0095906 0.51343 0.016013 2.9477 307.36 787.8 ± 68.3 406 92.47 0.0050470 0.51424 0.015057 1.5996 316.94 753.2 ± 74.8 407 92.07 0.0046780 0.53416 0.015054 1.4864 317.74 736.2 ± 54.7 408 91.22 0.0043500 0.54335 0.014809 1.3929 320.20 762.4 ± 79.0 409 89.39 0.0034291 0.51489 0.014694 1.1189 326.29 736.8 _ 28.9 501 87.74 0.0027632 0.51062 0.014879 0.91544 331.30 692.3 ± 36.0 502 89.30 0.0033005 0.51057 0.015277 1.0777 326.54 714.7 ± 53.7 503 97.29 0.0151460 0.52845 0.017340 4.5877 302.91 785.6 ± 57.8 504 91.41 0.0040769 0.51150 0.014884 1.3045 319.97 698.0 ± 45.6 505 89.57 0.0038014 0.53020 0.014863 1.2382 325.72 803.4 ± 57.5 506 82.33 0.0021395 0.52165 0.014837 0.74884 350.00 810.3 ± 96.3 507 75.18 0.0013408 0.50939 0.014134 0.50464 376.36 752.2 ± 25.0 508 78.42 0.0015370 0.50200 0.014692 0.55961 364.09 729.6 ± 34.4 509 85.19 0.0022572 0.48441 0.016824 0.76884 340.61 698.3 ± 57.2 510 92.02 0.0046661 0.51168 0.014970 1.4852 318.30 740.3 ± 75.3

J-value = 0.0035297 ± 0.0000341. Mean age = 746.7 ± 59.6 ka, STDEV = 37.8 ka.

No. Atmos. (%) 36Ar/39Ar 37Ar/39Ar 38Ar/39Ar 40Ar/39Ar 40Ar/36Ar Date (ka)

W-2 (Wenchang, China) 1401 32.08 0.00032272 0.52072 0.013202 0.20364 631.00 756.7 ± 88.0 1402 13.55 0.00020247 0.54105 0.013434 0.16932 836.26 774.2 ± 30.5 1403 9.53 0.00017516 0.52379 0.013595 0.15761 899.79 742.8 ± 54.1 1404 23.82 0.00025239 0.53200 0.013557 0.17352 687.48 702.7 ± 73.3 1405 16.03 0.00021426 0.53140 0.013696 0.17390 811.63 776.6 ± 55.7 1406 29.58 0.00029306 0.53547 0.013473 0.18506 631.49 701.4 ± 49.6 1407 9.77 0.00017621 0.52387 0.013665 0.15755 894.13 740.5 ± 36.1 1408 25.16 0.00026730 0.52272 0.013858 0.18606 696.06 750.0 ± 77.9 1409 11.78 0.00018499 0.52702 0.013482 0.15565 841.38 713.3 ± 44.6 1410 23.85 0.00025434 0.52239 0.013549 0.17877 702.89 727.9 ± 64.3 1411 24.64 0.00026454 0.53463 0.013541 0.18247 689.75 738.1 ± 44.5 1412 8.46 0.00016745 0.51747 0.013516 0.15267 911.72 722.8 ± 31.0 1413 21.97 0.00024212 0.51873 0.013300 0.17641 728.60 734.1 ± 42.4 1414 11.75 0.00018615 0.52641 0.013657 0.15929 855.69 734.1 ± 48.2

J-value = 0.0035297 ± 0.0000341. Mean age = 736.8 ± 55.5 ka, STDEV = 23.0 ka.

in core 17957-2 and ODP Hole 1144A as well as data small area in Laos has the greatest concentration of Muong from Glass and Pizzuto (1994) and Lee and Wei (2000), Nong-type layered tektites. All Australasian tektites may Glass (2003) predicted a source crater location for Aus- be derived from a single sedimentary formation with a tralasian strewn field near 15°N and 105°E which is 3° narrow range of stratigraphic ages close to ~170 Ma (Fig. farther north and 1° farther west than that predicted by 12). The field work of Sato (1992) shows that the limited Glass and Pizzuto (1994) and Lee and Wei (2000) but area near the southern part of the Thailand-Loas border closer to the location predicted by Schnetzler (1992). In proposed by Schnetzler (1992) as the potential impact site addition, Ford (1988) noted that the size of Muong Nong- has non-marine Jurassic exposures (Fig. 1) which is con- type tektites in Thailand decreased toward the west, and sistent with the hypothesis suggested by Blum et al. that Muong Nong-type tektites from Hainan Island, China (1992) based on Sr isotopic data. The geochemical data are small. Schnetzler and McHone (1996) noted that a discussed above support the hypothesis that the source

12 Y.-T. Lee et al. Table 5. (continued)

No. Atmos. (%) 36Ar/39Ar 37Ar/39Ar 38Ar/39Ar 40Ar/39Ar 40Ar/36Ar Date (ka)

KK-1 (Khon-Kaen, Thailand) 511 98.68 0.027340 0.41756 0.018185 8.1837 299.34 685.5 ± 77.5 512 98.78 0.032774 0.41111 0.019234 9.8017 299.07 759.1 ± 63.5 513 92.57 0.0051497 0.41798 0.014366 1.6386 318.19 761.9 ± 75.8 514 91.08 0.0042000 0.42543 0.013739 1.3561 322.88 753.8 ± 69.7 515 98.25 0.022384 0.42618 0.017002 6.7282 300.59 747.0 ± 92.4 601 90.81 0.0043560 0.41713 0.013682 1.4116 324.06 809.7 ± 74.0 602 88.05 0.0032771 0.42628 0.013444 1.0920 333.22 809.1 ± 58.8 603 91.33 0.0041593 0.41890 0.013517 1.3399 322.15 724.3 ± 54.5 604 92.01 0.0048600 0.41862 0.013739 1.5553 320.02 777.0 ± 70.6 605 94.93 0.0083177 0.41349 0.014548 2.5850 310.78 824.9 ± 63.6 606 93.97 0.0059652 0.42007 0.013947 1.8709 313.63 707.7 ± 79.2 607 96.61 0.011543 0.41690 0.014881 3.5267 305.53 754.7 ± 46.8 608 99.03 0.038451 0.42149 0.019824 11.470 298.30 704.4 ± 67.8

J-value = 0.0035297 ± 0.0000341. Mean age = 755.3 ± 69.7 ka, STDEV = 42.8 ka.

No. Atmos. (%) 36Ar/39Ar 37Ar/39Ar 38Ar/39Ar 40Ar/39Ar 40Ar/36Ar Date (ka)

M-01 (Maoming, China) 301 90.81 0.0043038 0.38773 0.015238 1.3969 324.58 800.4 ± 48.0 302 92.28 0.0049768 0.37834 0.015454 1.5915 319.78 768.2 ± 21.8 303 94.74 0.0067283 0.37793 0.016117 2.0972 311.70 692.9 ± 44.4 304 93.41 0.0056891 0.37774 0.015760 1.7980 316.05 742.9 ± 60.5 305 87.64 0.0029052 0.37928 0.015002 0.97562 335.82 745.2 ± 47.0 306 88.13 0.0031497 0.38458 0.015370 1.0519 333.97 773.5 ± 73.6 307 84.43 0.0022619 0.37784 0.015062 0.78660 347.77 751.4 ± 24.4 308 91.06 0.0044110 0.37966 0.015302 1.4287 323.89 797.0 ± 49.1 309 93.00 0.0053403 0.38245 0.015397 1.6945 317.31 742.4 ± 66.2 310 89.92 0.0035305 0.38065 0.015159 1.1569 327.70 723.9 ± 36.7 401 88.96 0.0031163 0.38292 0.015361 1.0314 330.96 704.7 ± 63.2 402 96.01 0.0090660 0.37594 0.016509 2.7895 307.69 701.3 ± 50.2 403 81.02 0.0016783 0.38197 0.015111 0.60532 360.66 697.1 ± 71.6

J-value = 0.0035297 ± 0.0000341. Mean age = 741.6 ± 52.9 ka, STDEV = 36.5 ka.

crater for Australasian strewn field appears to be located sandstone, siltstone, shale, mudstone and conglomerates in a limited area near the southern part of the Thailand- with minor intercalation of limestone are exposed (Ridd, Laos border. 1978; Sato, 1992). Regrettably, appropriate chemical data of target materials at the potential impact site at the Thai- land-Laos border are not available at present. In order to MIXING CALCULATIONS FOR APPROPRIATE TARGET establish a geochemical relationship between all the thirty MATERIAL OF TEKTITES FROM EAST ASIA tektites in this study and their parent materials, the present On the basis of previous studies (Koeberl, 1992; Blum authors selected a variety of target rocks in the literatures et al., 1992; Schnetzler, 1992) and the chemical data of (see Table 6) including average post-Archean Australian tektites analyzed in the present study, the authors suggest shale, average Phanerozoic sandstone and average Meso- that these tektites may represent the result of melting at a Cenozoic greywacke, and perform mixing calculations single site. The tektites of the present study are most likely based on the model proposed by Ho and Chen (1996) for to be derived as splashed melt from Jurassic sedimentary various amounts as well as combinations of these rocks. rocks forming the surface layer near the southern part of The results of the mixing model for all the major and trace the Thailand-Laos border where Jurassic non-marine elements are listed in Table 6.

Geochemical studies of tektites 13 Table 5. (continued)

No. Atmos. (%) 36Ar/39Ar 37Ar/39Ar 38Ar/39Ar 40Ar/39Ar 40Ar/36Ar Date (ka)

M-03 (Maoming, China) 2801 68.05 0.0010301 0.39690 0.014947 0.43204 419.41 865.4 ± 90.8 2802 15.00 0.00017409 0.39591 0.014404 0.17280 992.56 822.7 ± 69.4 2803 11.56 0.00015221 0.39670 0.014574 0.15916 1045.7 774.9 ± 52.6 2804 21.37 0.00021317 0.39879 0.014640 0.18284 857.70 814.0 ± 127.1 2805 7.36 0.00013248 0.39377 0.014561 0.15750 1188.8 801.5 ± 87.8 2901 23.53 0.00022858 0.38300 0.014462 0.19307 844.68 844.2 ± 110.5 2902 5.83 0.00012370 0.38677 0.014323 0.15576 1259.1 803.6 ± 158.0 2903 2.99 0.00011248 0.38949 0.014838 0.15889 1412.6 848.3 ± 64.0 2903 14.33 0.00016712 0.39157 0.014752 0.16746 1002.0 798.5 ± 64.0 2904 14.82 0.00017135 0.39002 0.014369 0.17205 1004.1 820.1 ± 68.2 2905 14.76 0.00016924 0.39179 0.014800 0.16748 989.64 794.5 ± 72.1 2906 4.13 0.00011354 0.37554 0.014749 0.15596 1373.6 819.5 ± 38.0 2907 14.33 0.00016712 0.39157 0.014752 0.16746 1002.0 798.5 ± 64.0 2907 6.50 0.00012690 0.38793 0.014622 0.15598 1229.1 799.4 ± 56.2

J-value = 0.00372265 ± 0.00002407. Mean age = 814.6 ka, STDEV = 24.4 ka.

Note J-value: Weighted mean of the results for twenty three fusions of irradiation standard LP-6 Biotite, having a K-Ar age of 127.7 ± 1.4 Ma (Odin et al., 1982). The date is obtained by using the following equation:

1  40 Ar∗  Dateln1 J , and =+ 39  λ  ArK  40 39 36 39 36 37 37 39 40 ∗ Ar/./.// Ar− 295 5 Ar Ar+ 295 5 Ar Ar Ar Ar 40 Ar ()m ()mCa()()m  Ar  39 = 39 37 37 39 −  39  ArK 1 − Ar// Ar Ar Ar  Ar  ()()Ca m K

where ( )Ca and ( )K = isotope ratios of argon extracted from irradiated calcium and potassium salts; ( )m = isotope ratio of argon extracted from irradiated unknown. Ð10 Ð1 Ð10 Ð1 Ð10 Ð1 Date (Ma) = the date calculated using the following decay constants: λε = 0.581 × 10 yr ; λβ = 4.962 × 10 yr ; λ = 5.543 × 10 yr ; 40K/K = 0.01167 atom % (Steiger and Jager, 1977). The quoted error is one standard deviation and does not include errors in the interference corrections.

The mixing model does not provide a perfect fit (Fig. Sm/Sc (avg. 0.43), Th/Sc (avg. 0.99) and REE contents 13), some major elements (e.g., Na2O and MgO) and trace of tektites studied are similar to the average upper conti- elements (e.g., Hf, Ta, Th, and U) contents do not agree nental crust. The La/Th ratio (avg. 2.84) and REE pat- exactly. Besides, the REE contents in the mixing model terns of East Asia tektites are similar to those of post- are generally lower than those of the tektites analyzed Archean sediments. Based on the results of the chemical (Table 6). However the model is useful in determining and isotopic data of this study, the lunar volcanic origin the most possible fit for the parent material of the tektites (Futrell, 2000) and multiple craters origin (Wasson, 1987, from East Asia. The best fit for the average composition 1991) may be excluded. The Ar-Ar ages of tektites of this of thirty tektites analyzed by the present study is a mix- study are consistent with the published ages of the ture of 61% greywacke, 32% sandstone and 7% shale. Australasian tektites which indicate that the Australasian tektites were derived from a single impact event. On the basis of Rb-Sr isotopic data, the tektites in this study were CONCLUSIONS derived from a single sedimentary formation with a Based on REE patterns the present authors excluded stratigraphic age close to 170 Ma (Jurassic). The poten- the lower crust and oceanic crust as the source rocks. The tial impact site may be located in a limited area near the trace elememt ratios Ba/Rb (avg. 3.74), Th/Sm (avg. 2.31), southern part of the Thailand-Laos border as suggested

14 Y.-T. Lee et al. Table 6. Average data for East Asia tektites (wt% and Koeberl of University of Vienna and Dr. B. P. Glass of Univer- ppm, for major and trace elements respectively) com- sity of Delaware for their penetrating reviews that lead to sig- pared with mixing model (M) proposed by Ho and Chen nificant improvement of the manuscript. The present study was (1996). The mixing model is constructed based on all supported by the National Science Council of the Republic of the major and trace elements analyzed China.

APAAS APSS AMCG M REFERENCES [wt%]

SiO 2 74.36 62.80 91.50 66.30 74.12 Barnes, V. E. (1961) Tektites. Sci. American 205 (5), 58Ð65. Al2O3 11.13 18.90 3.62 15.50 11.94 Bentley, C. R. (1979) No giant meteorite crater in Wilkes Land, MgO 2.59 2.20 0.45 2.00 1.52 Antarctic. J. Geophys. Res. 84, 5681Ð5682. FeO 4.99 6.50 1.13 6.20 4.60 Blum, J. D., Papanastassiou, D. A., Koeberl, C. and Wasserburg, CaO 2.14 1.30 0.31 3.20 2.14 G. J. (1992) Nd and Sr isotopic study of Australasian Na O 1.43 1.20 0.42 3.10 2.11 2 tektites: New constraints on the provenance and age of tar- K O 2.39 3.70 0.91 2.30 1.95 2 get material. Geochim. Cosmochim. Acta 56, 483Ð492. TiO 0.78 1.00 0.25 0.72 0.59 2 Burns, C. A. and Glass, B. P. (1989) Source region for the P2O5 0.04 0.16 0.02 0.14 0.10 Total 99.97 97.76 99.61 99.46 99.39 Australasian tektite strewn field. Meteoritics 24, 257. [ppm] Chapman, D. A. and Scheiber, L. C. (1969) Chemical investi- Sc 17.5 16.0 2.0 14.0 10.3 gation of Australasian tektites. J. Geophys. Res. 74, 6737Ð V 181 150 15 130 95 6794. Cr 79.7 110.0 30.0 70.0 60.0 Compston, W. and Chapman, D. R. (1969) Sr isotope pattern Co 17.4 23.0 2.5 15 11.6 within the Southeast Australasian strewn field. Geochim. Ni 14.3 55.0 8.0 30.0 24.7 Cosmochim. Acta 33, 1023Ð1036. Rb 115 160 25 100 80 Condie, K. C. (1993) Chemical composition and evolution of Sr 150 200 35 280 196 Y 47.4 27.0 6.8 28.0 21.1 the upper continental crust: Contrasting results from sur- Zr 418 210 105 145 137 face samples and shales. Chem. Geol. 104, 1Ð37. Nb 21.6 19.0 4.0 10.0 8.7 Englert, P., Pal, D. K., Tuniz, C., Moniot, R. K., Savin, W., Ba 430 650 150 650 490 Kruse, T. and Herzog, G. F. (1984) Manganese-53 and be- La 48.9 38.0 10.3 28.0 23.0 ryllium-10 contents of tektites (abs.). Lunar Planet. Sci. 15, Ce 96.5 80.0 22.3 61.0 49.9 250Ð251. Nd 39.2 32.0 8.4 26.0 20.8 Evensen, M. N., Hamilton, P. J. and O’Nions, R. K. (1978) Rare- Sm 7.5 5.6 1.6 4.9 3.9 earth abundances in chondritic meteorites. Geochim. Eu 0.98 1.10 0.34 0.90 0.73 Cosmochim. Acta 42, 1199Ð1212. Gd 5.72 4.70 1.44 4.34 3.44 Ford, R. J. (1988) An empirical model for the Australasian Tb 0.88 0.77 0.21 0.66 0.52 Yb 3.03 2.80 0.61 2.20 1.73 tektite strewn field. Australasian J. Earth Sci. 35, 483Ð490. Lu 0.50 0.43 0.11 0.38 0.30 Fuhrmann, U., Lippolt, H. J. and Hess, J. C. (1987) Examina- 40 39 Hf 8.31 5.00 3.10 3.90 3.72 tion of some proposed K-Ar standards: Ar/ Ar analyses Ta 1.3 1.2 0.3 0.9 0.7 and conventional K-Ar dating. Chem. Geol., Isot. Geosci. Th 17.2 14.6 4.0 8.5 7.5 Sect. 66, 41Ð51. U 2.65 3.10 1.10 1.80 1.67 Futrell, D. S. (2000) Tektite controversy. Meteorite 6, 36Ð37. A, average of 30 East Asia tektites (this study); PAAS, average post- Gentner, W., Kleinmann, B. and Wagner, G. A. (1967) New K- Archean Australian Shales (Condie, 1993); APSS, average Phanero- Ar and fission track ages for impact glasses and tektites. zoic sandstones (Condie, 1993); AMCG, average Meso-Cenozic Earth Planet. Sci. Lett. 2. 83Ð86. greywackes (Condie, 1993); M, 7% PAAS + 32% APSS + 61% AMCG. Glass, B. P. (1979) Zhamanshin crater, a possible source of Australasian tektites. Geology 7, 351Ð353. Glass, B. P. (1990) Tektites and microtektites: Key facts and inferences. Tectonophysics 171, 393Ð404. Glass, B. P. (1993) Geographic variations in abundance of Aus- by Schnetzler (1992). Mixing calculations for various tralasian microtektites: Implications concerning the loca- amounts and combinations of target rocks indicate that tion and size of the source crater (abs.). Meteoritics 28, 354. the best fit for the average composition of thirty tektites Glass, B. P. (2000) Tektite controversy (reply to Darryl S. Futrell’s comments). Meteorite 6, 38Ð40. in this study is a mixture of 61% greywacke, 32% sand- Glass, B. P. (2003) Australasian microtektites in the South China stone and 7% shale. Sea: Implications regarding the location and size of the source crater. Lunar and Planetary Science Conference 34. Acknowledgments—We thank Dr. C. H. Lo of National Glass, B. P. and Pizzuto, J. E. (1994) Geographical variation in Taiwan University for technical assistance in Ar-Ar dating. We Australasian microtektite concentrations: Implications con- are grateful to Mr. R. J. Lin and Ms. Y. Y. Su who assisted with cerning the location and size of the source crater. J. Geophys. the chemical analyses of some tektite samples. We thank Dr. C. Res. 99, 19075Ð19081.

Geochemical studies of tektites 15 Govindaraju, K. (1994) Compilation of working values and 73Ð78. sample descriptions for 383 geostandards. Geostand. Pal, D. K., Tuniz, C., Moniot, R. K., Kruse, T. H. and Herzog, Newlett. 18, 1Ð158. G. F. (1982) Beryllium-10 in Australasian tektite: Evidence Hartung, J. B. and Rivolo, A. R. (1979) A possible source in for a sedimentary precursor. Science 218, 787Ð789. Cambodia for Australasian tektites. Meteoritics 14, 153Ð Poag, C. W., Powars, D. S., Poppe, L. J. and Mixon, R. B. (1994) 159. Meteoroid mayhem in Ole Virginny: Source of the North Ho, K. S. and Chen, J. C. (1996) Geochemistry and origin of American tektite strewn field. Geology 22, 691Ð694. tektites from the Penglei area, Hainan province, southern Ridd, M. F. (1978) The Mesozoic: Thailand. The Phanerozoic China. J. SE. Asian Earth Sci. 13, 61Ð72. Geology of the World II (Moullade, M. and Nairn, A. E. M., Hofmann, A. W. (1988) Chemical differentiation of the Earth: eds.), 145Ð161, Elsevier. The relationship between mantle, continental crust, and oce- Rudnick, R. L. (1992) Xenoliths-sample of the lower continen- anic crust. Earth Planet. Sci. Lett. 90, 297Ð314. tal crust. Continental Lower Crust (Fountain, D. M., Izett, G. A. and Obradovich, J. D. (1992) Laser-fusion 40Ar/ Arculus, R. and Kay, R. W., eds.), 269Ð316, Elsevier. 39Ar ages of Australasian tektites (abstract). Lunar Planet. Sato, T. (1992) Regional geology and stratigraphy: Southeast Sci. 23, 593Ð594. Asia and Japan. The Jurassic of the Circum-Pacific Koeberl, C. (1986) Geochemistry of tektites and impact glasses: (Westermann, G. E. G., ed.), 194Ð213, Cambridge Univer- An overview. Ann. Rev. Earth Planet. Sci. 14, 325Ð350. sity Press. Koeberl, C. (1990) The geochemistry of tektites: An overview. Schmidt, G., Zhou, L. and Wasson, J. T. (1993) Iridium anomaly Tectonophysics 171, 405Ð422. associated with the Australasian tektite-producing impact: Koeberl, C. (1992) Geochemistry and origin of Muong Nong- Masses of the impactor and of the Australasian tektites. type tektites. Geochim. Cosmochim. Acta 56, 1033Ð1064. Geochim. Cosmochim. Acta 57, 4851Ð4859. Koeberl, C. (1994) Tektite origin by hypervelocity asteroidal Schnetzler, C. C. (1992) Mechanism of Muong Nong-type or cometary impact: Target rocks, source craters, and mecha- tektite formation and speculation on the source of Australa- nisms. Basins of the Rio Grande Rift: Structure, sian tektites. Meteoritics 27, 154Ð165. Stratigraphy, and Tectonic Setting (Keller, G. R. and Cather, Schnetzler, C. C. and McHone, J. F. (1996) Source of Australa- S. M., eds.), Geol. Soc. Am. Special Paper 293, 133Ð151. sian tektites: Investigating possible sites in Laos. Meteoritics Koeberl, C. and Fredriksson, K. (1986) Impact glasses from Planet. Sci. 31, 73Ð76. the Zhamanshin crater, USSR: Chemical composition and Schwarz, W. H. and Lippolt, H. J. (2002) Coeval argon-40/ar- discussion of origin. Earth Planet. Sci. Lett. 78, 80Ð88. gon-39 ages of moldavites from the Bohemian and Lusatian Koeberl, C., Poag, C. W., Reimold, W. U. and Brandt, D. (1996) strewn fields. Meteoritics Planet. Sci. 37, 1757Ð1763. Impact origin of the Chesapeake Bay structure and the Shaw, H. F. and Wasserburg, G. J. (1982) Age and provenance source of the North American tektites. Science 271, 1263Ð of the target materials for tektites and possible impactites 1266. as inferred from Sm-Nd and Rb-Sr systematic. Earth Planet. Koeberl, C., Bottomley, R., Glass, B. P. and Storzer, D. (1997) Sci. Lett. 60, 155Ð177. Geochemistry and age of Ivory Coast tektites and Stauffer, P. H. (1978) Anatomy of the Australasian tektite microtektites. Geochim. Cosmochim. Acta 61, 1745Ð1772. strewn-field and the probable site of its source crater. Pro- Lacroix, A. (1935) Les tectites sans formes figurees de ceedings of the 3rd Regional Conference on Geology and I’Indochine. Compt. Rend. Acad. Sci. Paris 200, 2129Ð2132. Mineral Resources of Southeast Asia (Nutalaya, P., ed.), Lee, M. Y. and Wei, K. Y. (2000) Australasian microtektites in 285Ð289. the South China Sea and the western Philippine Sea: Impli- Steiger, R. H. and Jager, E. (1977) Subcommission on cations for age, size and location of the impact crater. geochronology: Convention on the use of decay constant in Meteoritics Planet. Sci. 35, 1151Ð1155. geo- and cosmochronology. Earth Planet. Sci. Lett. 36, 359Ð Lo, C. H. and Lee, C. Y. (1994) 40Ar-39Ar method of K-Ar age 362. determination of geological samples using Tsing-Hua Open Storzer, D. and Wagner, G. A. (1971) Fission track ages of North Pool Reactor (THOR). J. Geol. Soc. China 37(1), 1Ð22. American tektites. Earth Planet. Sci. Lett. 10, 435Ð444. Matsubara, K., Matsuda, J. I. and Koeberl, C. (1991) Noble Taylor, S. R. (1969) Criteria for the source of australites. Chem. gases and K-Ar ages in Aouelloul, Zhamanshin, and Libyan Geol. 4, 451Ð459. Desert impact glasses. Geochim. Cosmochim. Acta 55, Taylor, S. R. (1973) Tektites: A post-Apollo view. Earth Sci. 2951Ð2955. Rev. 9, 101Ð123. McLennan, S. M. and Taylor, S. R. (1980) Th and U in sedi- Taylor, S. R. and McLennan, S. M. (1979) Chemical relation- mentary rocks: Crustal evolution and sedimentary recycling. ships among irghizites, Zhamanshinites, Australasian Nature 285, 621Ð624. tektites and Henbury impact glasses. Geochim. Cosmochim. Odin, G. S. (1982) Interlaboratory standards for dating potas- Acta 43, 1551Ð1565. sium-argon age determinations. Numerical Dating in Taylor, S. R. and McLennan, S. M. (1985) The Continental Stratigraphy (Odin, G. S., ed.), 123Ð149, Wiley, Chiches- Crust: Its Composition and Evolution. Blackwell, London, ter. 312 pp. O’Keefe, J. A. (1976) Tektites and Their Origin. Elsevier, Taylor, S. R., McLennan, S. M. and McCulloch, M. T. (1983) Amsterdam, 254 pp. Geochemistry of loess, continental crust composition and O’Keefe, J. A. (1994) The origin of tektites. Meteoritics 29, crustal model ages. Geochim. Cosmochim. Acta 47, 1897Ð

16 Y.-T. Lee et al. 1905. rials. Geochim. Cosmochim. Acta 33, 1045Ð1055. Tera, F., Brown, L., Klein, J., Middleton, R. and Mason, B. Wasson, J. T. (1987) A multiple-impact origin of southeast Asian (1983) Beryllium-10 and aluminium-26 in tektites. tektites (abs.). Lunar Planet. Sci. 18, 1062Ð1063. Meteoritics 18, 405Ð406. Wasson, J. T. (1991) Layered tektites: A multiple impact origin Wampler, J. M., Smith, D. M. and Cameron, A. E. (1969) Iso- for the Australasian tektites. Earth Planet. Sci. Lett. 102, topic comparison of lead in tektites with lead in earth mate- 95Ð109.

Geochemical studies of tektites 17