Show and Tell

Old Pages


Carolina Bays?







Contact George



Click here to access the FREE PAPER pdf from pnas




Proceedings of the National Academy of Sciences

Published online before print September 27, 2007
Proc. Natl. Acad. Sci. USA, 10.1073/pnas.0706977104
Evidence for an extraterrestrial impact 12,900 years ago that contributed to the megafaunal extinctions and the Younger Dryas cooling

( comet | iridium | micrometeorites | nanodiamonds | spherules )

R. B. Firestone a,b, A. West c, J. P. Kennett d, L. Becker e, T. E. Bunch f, Z. S. Revay g, P. H. Schultz h, T. Belgya g, D. J. Kennett i, J. M. Erlandson i, O. J. Dickenson j, A. C. Goodyear k, R. S. Harris h, G. A. Howard l, J. B. Kloosterman m, P. Lechler n, P. A. Mayewski o, J. Montgomery j, R. Poreda p, T. Darrah p, S. S. Que Hee q, A. R. Smith a, A. Stich r, W. Topping s, J. H. Wittke f, and W. S. Wolbach r

aLawrence Berkeley National Laboratory, Berkeley, CA 94720; cGeoScience Consulting, Dewey, AZ 86327; dDepartment of Earth Sciences and eInstitute of Crustal Studies, University of California, Santa Barbara, CA 93106; fNorthern Arizona University, Flagstaff, AZ 86011; gInstitute for Isotope and Surface Chemistry, H-1525, Budapest, Hungary; hDepartment of Geological Sciences, Brown University, Providence, RI 02912; iDepartment of Anthropology and Museum of Natural and Cultural History, University of Oregon, Eugene, OR 97403; jEastern New Mexico University, Portales, NM 88130; kSouth Carolina Institute of Archaeology and Anthropology, University of South Carolina, Columbia, SC 29208; lRestoration Systems, LLC, Raleigh, NC 27604; mRozenstraat 85, 1018 NN, Amsterdam, The Netherlands; nBureau of Mines and Geology, University of Nevada, Reno, NV 89557; oClimate Change Institute, University of Maine, Orono, ME 04469; pUniversity of Rochester, Rochester, NY 14627; qDepartment of Environmental Health Sciences, University of California, Los Angeles, CA 90095; sP.O. Box 141, Irons, MI 49644; and rDepartment of Chemistry, DePaul University, Chicago, IL 60614

Communicated by Steven M. Stanley, University of Hawaii at Manoa, Honolulu, HI, July 26, 2007 (received for review March 13, 2007)

A carbon-rich black layer, dating to {approx}12.9 ka, has been previously identified at {approx}50 Clovis-age sites across North America and appears contemporaneous with the abrupt onset of Younger Dryas (YD) cooling. The in situ bones of extinct Pleistocene megafauna, along with Clovis tool assemblages, occur below this black layer but not within or above it. Causes for the extinctions, YD cooling, and termination of Clovis culture have long been controversial. In this paper, we provide evidence for an extraterrestrial (ET) impact event at {cong}12.9 ka, which we hypothesize caused abrupt environmental changes that contributed to YD cooling, major ecological reorganization, broad-scale extinctions, and rapid human behavioral shifts at the end of the Clovis Period. Clovis-age sites in North American are overlain by a thin, discrete layer with varying peak abundances of (i) magnetic grains with iridium, (ii) magnetic microspherules, (iii) charcoal, (iv) soot, (v) carbon spherules, (vi) glass-like carbon containing nanodiamonds, and (vii) fullerenes with ET helium, all of which are evidence for an ET impact and associated biomass burning at {approx}12.9 ka. This layer also extends throughout at least 15 Carolina Bays, which are unique, elliptical depressions, oriented to the northwest across the Atlantic Coastal Plain. We propose that one or more large, low-density ET objects exploded over northern North America, partially destabilizing the Laurentide Ice Sheet and triggering YD cooling. The shock wave, thermal pulse, and event-related environmental effects (e.g., extensive biomass burning and food limitations) contributed to end-Pleistocene megafaunal extinctions and adaptive shifts among PaleoAmericans in North America.

Author contributions: R.B.F., A.W., J.P.K., L.B., and W.T. designed research; R.B.F., A.W., J.P.K., L.B., T.E.B., Z.S.R., P.H.S., D.J.K., J.M.E., O.J.D., A.C.G., R.S.H., G.A.H., J.B.K., P.L., P.A.M., J.M., R.P., T.D., S.S.Q.H., A.R.S., A.S., W.T., J.H.W., and W.S.W. performed research; R.B.F., A.W., J.P.K., L.B., T.E.B., Z.S.R., T.B., D.J.K., O.J.D., A.C.G., G.A.H., J.B.K., P.L., J.M., R.P., S.S.Q.H., W.T., J.H.W., and W.S.W. contributed new reagents/analytic tools; R.B.F., A.W., J.P.K., L.B., T.E.B., Z.S.R., P.H.S., D.J.K., J.M.E., R.S.H., G.A.H., P.A.M., R.P., T.D., S.S.Q.H., A.R.S., A.S., W.T., J.H.W., and W.S.W. analyzed data; and R.B.F., A.W., J.P.K., and P.H.S. wrote the paper.

The authors declare no conflict of interest.

Freely available online through the PNAS open access option.

bTo whom correspondence should be addressed.

R. B. Firestone, E-mail: rbfirestone@lbl.gov



Supporting material below

formal paper here

Fig. 5. The dark line shown above is the black mat (12.9 ka) along the arroyo wall of the Murray Springs Clovis site in Arizona. The YDB markers, including magnetic grains and microspherules, iridium, soot, and fullerenes with ET helium, are present in the few centimeters just below the black mat at the top of the underlying sediment. This lithologic break represents the surface at the end of the Clovis period before the formation of the black mat. Clovis artifacts, a fire pit, and an almost fully articulated skeleton of an adult mammoth were recovered at Murray Springs with the black mat draped conformably over them. Excavations by Vance Haynes, Jr., and colleagues also revealed hundreds of mammoth footprints in the sand infilled by black mat sediments. These footprints and the mammoth skeleton appear to have been preserved by rapid burial after the YDB event (1). No in situ Clovis points and extinct megafaunal remains have been recovered from in or above the black mat, indicating that the mammoths (except in isolated cases) and Clovis hunting technology disappeared simultaneously.

1. Haynes CV, Jr (1987) Centennial Field Guide Volume 1: Cordilleran Section of the Geological Society of America (Geolog Soc Am, Boulder, CO), Vol. 1, pp 23-28.


Fig. 6. Clovis and the Younger Dryas. Haynes, in Taylor et al. (1), correlated the end of Clovis cultural adaptations with the onset of Younger Dryas cooling and provided end-Clovis 14C dates that have been calibrated to 12.92 ka for Murray Springs and 12.98 ka for Blackwater Draw, two of the sites we analyzed. This graph displays a corresponding date of 12.9 ka for the onset of the YD in Greenland GISP2 ice core data based on paleotemperature analyses (ref. 2, in red) and changes in methane concentrations (ref. 3, in blue). The onset of the YD was marked by a dramatic 8°C drop in Greenland temperature in <150 years with an associated abrupt decrease in atmospheric methane concentrations. We propose that these climatic changes were triggered by the YD event at ≈12.9 ka.

1. Taylor RE, Haynes CV, Stuiver M (1996) Antiquity 70:515-525.

2. Alley RB (2000) Quaternary Sci Rev 19:213-226.

3. Brook EJ, Harder S, Severinghaus J, Steig EJ, Sucher CM (2000) Cycles 14(2):559-572.


Fig. 7. Aerial photo (U.S. Geological Survey) of a cluster of elliptical and often overlapping Carolina Bays with raised rims in Bladen County, North Carolina. The Bays have been contrast-enhanced and selectively darkened for greater clarity. The largest Bays are several kilometers in length, and the overlapping cluster of them in the center is ≈8 km long. Previous researchers have proposed that the Bays are impact-related features.


Fig. 8. Lommel (1) is in northern Belgium, near the border with the Netherlands. At 12.94 ka (2), this site was a large late Glacial sand ridge covered by open forest at the northern edge of a marsh. More than 50 archaeological sites in this area indicate frequent visits by the late Magdalenians, hunter-gatherers who were contemporaries of the Clovis culture in North America. Throughout the Bölling-Allerod, eolian sediments known as the Coversands blanketed the Lommel area. Then, just before the Younger Dryas began, a thin layer of bleached sand was deposited and, in turn, was covered by the dark layer marked "YDB" above. That stratum is called the Usselo Horizon and is composed of fine to medium quartz sands rich in charcoal. The dark Usselo Horizon is stratigraphically equivalent to the YDB layer and contains a similar assemblage of impact markers (magnetic grains, magnetic microspherules, iridium, charcoal, and glass-like carbon). The magnetic grains have a high concentration of Ir (117 ppb), which is the highest value measured for all sites yet analyzed. On the other hand, YDB bulk sediment analyses reveal Ir values below the detection limit of 0.5 ppb, suggesting that the Ir carrier is in the magnetic grain fraction. The abundant charcoal in this black layer suggests widespread biomass burning. A similar layer of charcoal, found at many other sites in Europe, including the Netherlands (3), Great Britain, France, Germany, Denmark, and Poland (4), also dates to the onset of the Younger Dryas (12.9 ka) and, hence, correlates with the YDB layer in North America. [Reproduced with permission from Marc De Bie (Copyright 2004).]

1. Hoek WZ (1997) Veget Hist Archaeobot 6:197-213.

2. Van Geel B, Coope GR, Vander Hammen T (1989) Rev Paleont Palyn 60:25-129.

3. Hoek WZ (1997) Veget Hist Archaeobot 6:197-213.

4. Kloosterman JB (1976) Catastroph Geol 1(2):57-58.

Fig. 9. Research sites with calibrated YDB ages, including Lommel, Belgium, shown in Inset. High-Ir sites are shown in green. For the Bays, three of five sediment analyses revealed detectable Ir values, although radiocarbon ages of the Bays are inconsistent. Sediments from sites with no detectable Ir values (<0.5 ppb) are shown in brown. Sites with black mats are marked with inverted triangles. The approximate extent of the North American ice sheets at 12.9 ka is shown in blue-green, which is consistent with our observations that all sites were ice-free at the time of the YD event.


Fig. 10. SEM photomicrographs of mostly individual particles of submicrometer-sized soot (shown on filter paper at yellow arrows), measured at 1,969 ± 167 ppm from Blackville Bay T13 (Left), and measured at 21 ± 7 ppm from Murray Springs (Right). The soot levels and morphology from both sites are similar to those from the K/T (1). Only two of eight sites tested exhibited soot, perhaps because of unfavorable conditions for preservation at some sites. Soot was identified using SEM imaging and quantified by particle size analysis and weighing (2).

1. Wolbach WS, Gilmour I, Anders E, Orth CJ, Brooks RR (1988) Nature 334:665-9.

2. Wolbach WS, Anders E (1989) Geochim Cosmochim Acta 53:1637-1647.

Fig. 11. A 13C NMR spectrum of glass-like carbon from Carolina Bay M33 in Myrtle Beach, South Carolina. This was produced on a Varian Unity-200 NMR spectrometer operating at 50.2 MHz and equipped with a Doty Scientific 7-mm Supersonic MAS probe. Spinning speeds of 6.5 kHz were used and a variable-amplitude, cross-polarization pulse sequence was used with recycle delays of 1 s and a contact time of 1 ms. The aliphatic carbon appears centered at 38 ppm, which is typical of peaks representative of nanodiamonds, where small diamond domains are formed in compressed aromatic/graphitic materials. Of the ≈9-10% aliphatic carbon, the inferred nanodiamond component is estimated to represent ≈3% total carbon.

Fig. 12. Deposition rates [calculated by MG ´ D ´%A´ p/100, where MG (measured in mg/g) is magnetic grain concentration (Table 1), D (in cm) is the YDB layer (Table 1), %A is percent mineral abundance (Table 2), and p (in g/cm3) is the average mineral density.] for magnetic microspherules, magnetic grains, and their principal components at YDB sites ordered by distance from the Gainey, MI, site (upper scale). Microspherules, magnetic grains, magnetite, silicates,, and water content all dramatically peak at Gainey, suggesting that they are terrestrial products of a nearby impact. Ilmenite/rutile concentrations peak at Topper and are higher than Gainey at all sites, suggesting that they are high-velocity ejecta from an impact. Because magnetic grains at Wally's Beach where recovered from inside an extinct Pleistocene horse skull and may not be representative of the sediment, magnetic grain concentrations there are normalized to those at the nearby Chobot site.


Fig. 13. Mammoth bone found with Clovis artifacts (from the Blackwater Draw collection). This bone is stained yellow (arrow) and is highly radioactive (3,000 ppm U) only on the upper side that was just below the black mat. Bones found above or deeper below the black mat are neither stained nor highly radioactive. INAA analysis determined a high U concentration (58 ppm) in YDB sediment at Blackwater Draw, which is ≈10 times the concentration above or below. High U content on fossil bones is due to well known diagenetic processes (1) as confirmed by the corresponding low Th content (<1 ppm) on the stained bone surface. During breakdown of organic material under anoxic conditions, bone beds also may precipitate phosphatic minerals (2), which in turn scavenge and concentrate U. If so, the U enrichment on the bones and in the YDB sediment may have been enhanced by the abundance of bones and other Ca sources in the extinction layer. High levels of radioactivity may, therefore, be potentially useful as an additional diagnostic marker of the YDB layer.

1. Hedges REM (2002) Archaeometry 44(3):319-328.

2. Purnachandra Rao V, Naqvi SWA, Dileep Kumar M, Cardinal D, Michard A, Borole DV, Jacobs E, Natarajan R (2000) Sedimentology 47(5):945-960.

Fig. 14. (Left) Radioactivity profiles measured with a Geiger counter at Blackwater Draw and Murray Springs. (Right) Radioactivity in bone fragments from Blackwater Draw sediments (1) are compared with U and Th concentrations from Blackwater Draw sediment. Radioactivity peaks in both sediment and bone fragments in the YDB due to high concentrations of U.

1. Fitting J (1963) in Studies in the Natural Radioactivity of Prehistoric Materials, eds Jelinek A, Fitting JE (Univ of Michigan, Ann Arbor), pp 66-66.


Fig. 15. Sediment concentrations for U, Th, Hf, Sc, and Sm peak in the K/T boundary at Gubbio, Italy (A) (1), and the late Eocene Chesapeake Bay impact (≈36 Ma) at Massignano, Italy (2), which produced one of the largest known tektite strewnfields (B). (C and D) Radioactive element concentrations also peak in the YDB at Blackwater Draw, NM, (C) and Lake Hind, Manitoba, Canada (D). At Blackwater Draw, the uranium increase as determined by INAA is especially large (58 ppm) and yielded the most radioactive sediment analyzed in the study (SI Fig. 14). Concentrations (in ppm) are shown on a log scale, and depth (in cm) is centered on the YDB layer. Ir, Ni, and numerous other elements also peak at the YDB layer (presented in the main text) and are considered to have resulted from impact processes.

1. Alvarez LW, Alvarez W, Asaro F, Michel HV (1980) Science 208:1095-1108.

2. Bodiselitsch B, Montanari A, Koeberl C, Coccioni R (2004) Earth Planet Sci Lett 223:283-302.


Fig. 16. Zircon (ZrSiO4) is one of several heavy minerals potentially enriched with U and Th that can be concentrated to form a radioactive layer. (Left) We analyzed sediment samples at the Topper site for Zr (red arrow), the major constituent of zircon, and found evidence for a minor increase in zircon abundance in the YBD at Topper. When normalized to crustal values, U (purple arrow), Th, and Hf concentrations changed in direct relationship to the abundance of Zr, suggesting that zircon may account for some of the increased radioactivity. (Right) In contrast, at Daisy Cave, U decreased relative to zircon, indicating a negative correlation to sediment radioactivity.

Fig. 17. Ti may appear in the heavy minerals ilmenite, rutile, and titanite. (Left) At Topper, the presence of sedimentary Ti (red arrow) correlates well with higher sedimentary levels of U (purple arrow) and Th. (Right) However, at Daisy Cave, these relationships were negative, as with zircon. In summary, heavy mineral concentrations tested do not correlate well with an increase in sediment radioactivity at Daisy Cave but do so at Topper, where the formation of lag deposits may have been influenced by the impact. Heavy minerals may be concentrated through impact-related processes such as (i) high-velocity winds associated with the shockwave; (ii) heavy rains and flooding following the impact; and (iii) selective dissolution of sediment by acidic conditions due to fallout and acid rain. However, it is unlikely that lag deposits are typical of the YDB, because these sediment sequences appear to be relatively continuous. Furthermore, such deposits would have concentrated interplanetary dust particles (IDPs), and they would be present in the magnetic fractions isolated from bulk sediments at BWD and Murray Springs. However, these two sites do not show high 3He/4He ratios in the magnetic fraction, such as would be present if the lag deposits had concentrated the IDPs, nor does the He in the bulk sediment suggest any such concentration at the boundary. Only the fullerenes concentrate ET He, which is inconsistent with lag deposits and consistent with an impact event at the YDB.

SI Text

Research Sites. Murray Springs. Near Sierra Vista, AZ, Murray Springs is one of several local Clovis mammoth kill-sites associated with a chain of end-Pleistocene ponds at 12.9 ka. Sediments from the YDB layer are mostly fine to coarse fluvial or lacustrine sand. A distinctive black mat, most likely of algal origin, drapes conformably over the bones of butchered mammoths, and a thin layer (<2 cm) that contains YDB markers lies at the base of the black mat and immediately overlies the bones (1). The upper surfaces of some Clovis-butchered mammoth bones, which were in direct contact with the YDB and the black mat, exhibit slightly higher radioactivity and magnetic susceptibility than the lower surfaces.

Blackwater Draw. Blackwater Draw, NM, is southwest of the town of Clovis, which gave its name to the type of projectile points first found there. It was a PaleoAmerican hunting site on the bank of a spring-fed waterhole, where the black mat was found draped over bones of butchered mammoths and Clovis artifacts. YDB markers are concentrated in a ≈2-cm layer of fine-grained fluvial or lacustrine sediment that lies at the base of the black mat in the uppermost stratigraphic horizon containing in situ mammal bones and Clovis artifacts. The upper surfaces of some mammal bones were in direct contact with the YDB or the black mat and exhibited very high levels of radioactivity. We sampled a 2-m stratigraphic sequence spanning the YDB down into the deep gravels that date to >40 ka and possibly to 1.6 Ma (2). ET markers peaked only in the YDB.

Gainey. North of Detroit, MI, Gainey was a PaleoAmerican campsite located tens of kilometers from the southern margin of the Laurentide Ice Sheet at 12.9 ka. Sediments containing YDB markers are mostly fine alluvial sand and glacial silt. The Gainey site has been closed and hence inaccessible for many years, and only archived samples from the ≈5-cm YDB layer were available for analysis. No black mat was observed.

Wally's Beach. At St. Mary Reservoir, southwestern AB, Canada, Wally's Beach was a stream-fed valley that, at 12.9 ka, supported many species of now-extinct megafauna, including mammoths, camels, and horses. Hundreds of their footprints were found there during prior excavations. A sediment sample of fine-grained and silty alluvium was provided to us by Brian Kooyman from the brain cavity of a horse skull found in the YDB layer amidst Clovis points that tested positive for horse protein, providing some of the first evidence that Clovis peoples hunted horses (3).

Topper. Topper is located on a high bank of the Savannah River near Allendale, SC, and was a Clovis-age flint quarry containing thousands of artifacts. Sediments are eolian, fluvial, colluvial, and alluvial in origin and are comprised mostly of coarse to medium quartz sand. YDB markers occur within a ≈5-cm interval immediately in and above a distinct layer of Clovis artifacts. Lower sediments in the sequence have been dated to >55 ka (4), and no ET markers appear in the stratigraphic sections above or below the YDB. There is no black mat at this site.

At a new excavation, we used the neodymium magnet and a magnetic susceptibility meter to help identify the YD layer based on the high iron content. Shortly afterward, the excavators recovered part of a Clovis point immediately beneath the YD layer, illustrating the usefulness of the YDB markers for locating the Clovis horizon in new locations.

Chobot. Chobot is Southwest of Edmonton, AB, Canada. In Clovis times, it was located along the shore of a proglacial lake, where a supply of quality flint attracted hunter-gatherers. The presence of Clovis artifacts (5) dates this level to an interval of ≈200 yr ending at 12,925 cal B.P. (6). The Clovis level is capped by the YDB layer, above which there is a black mat similar to other sites. The YDB sediment samples are mostly fine-grained and colluvial.

Daisy Cave. A cave/rockshelter on San Miguel Island, Daisy Cave is one of the Channel Islands off the Southern California coast. This cave does not appear to have been occupied until ≈11.5 ka, but a Clovis-age human skeleton was found on nearby Santa Rosa Island, demonstrating that the PaleoAmericans had boats capable of reaching the islands. Several markers were found, but others, including Ir, were not found, possibly because the protected cave shelter prevented accretion. The sediment with YDB markers dates to ≈13.09 ka (7) and varies from fine sand to silt.

Lake Hind. In MN, Canada, Lake Hind was an end-Pleistocene proglacial lake. Various analyses by Boyd et al. (8) show that prior to 12.76 ka, the ice dams on the lake failed catastrophically as part of a regional pattern of glacial lake drainages. In this study, we confirmed with calibrated radiocarbon dating that the drainage took place at ≈12.76 ka (UCIAMS 29317). At the YDB, the failure rapidly transformed the lake from deep to shallow water, as shown by pollen analysis and the start of peat accumulation. The sample sediments are fine-grained lacustrine silt and peat.

Morley. Morley is a nonarchaeological site west of Calgary in AB, Canada. The site is on a raised drumlin, a subglacial erosional landform that formed at the end of the Pleistocene during deglaciation (9). The largest drumlin field near Ontario (5,000 km2) contains 3,000 drumlins that date to shortly after 13 ka, and the age of the Morley drumlin field appears to be similar. Later, the ice sheet melted away leaving atop the drumlin glacial debris containing numerous YDB markers. Samples are mostly gravel grading down through coarse and medium sand.

Lommel. Lommel is described in SI Fig. 8.

Carolina Bays. The Carolina Bays are a group of »500,000 highly elliptical and often overlapping depressions scattered throughout the Atlantic Coastal Plain from New Jersey to Alabama (see SI Fig. 7). They range from ≈50 m to ≈10 km in length (10) and are up to ≈15 m deep with their parallel long axes oriented predominately to the northwest. The Bays have poorly stratified, sandy, elevated rims (up to 7 m) that often are higher to the southeast. All of the Bay rims examined were found to have, throughout their entire 1.5- to 5-m sandy rims, a typical assemblage of YDB markers (magnetic grains, magnetic microspherules, Ir, charcoal, soot, glass-like carbon, nanodiamonds, carbon spherules, and fullerenes with 3He). In Howard Bay, markers were concentrated throughout the rim, as well as in a discrete layer (15 cm thick) located 4 m deep at the base of the basin fill and containing peaks in magnetic microspherules and magnetic grains that are enriched in Ir (15 ppb), along with peaks in charcoal, carbon spherules, and glass-like carbon. In two Bay-lakes, Mattamuskeet and Phelps, glass-like carbon and peaks in magnetic grains (16-17 g/kg) were found ≈4 m below the water surface and 3 m deep in sediment that is younger than a marine shell hash that dates to the ocean highstand of the previous interglacial.

Modern Fires. Four recent modern sites were surface-sampled. Two were taken from forest underbrush fires in North Carolina that burned near Holly Grove in 2006 and Ft. Bragg in 2007. Trees mainly were yellow pine mixed with oak. There was no evidence of carbon spherules and only limited evidence of glass-like carbon, which usually was fused onto much larger pieces of charcoal. The glass-like carbon did not form on oak charcoal, being visible only on pine charcoal, where it appears to have formed by combustion of highly flammable pine resin.

Two surface samples also were taken from recent modern fires in Arizona; they were the Walker fire, which was a forest underbrush fire in 2007 and the Indian Creek Fire near Prescott in 2002, which was an intense crown fire. Trees mainly were Ponderosa pine and other species of yellow pine. Only the crown fire produced carbon spherules, which were abundant (≈200 per kg of surface sediment) and appeared indistinguishable from those at Clovis sample sites. Both sites produced glass-like carbon fused onto pine charcoal.

Methods. Separation of the magnetic fraction from sediment. Initially, we used the magnet for in situ field testing to help locate the peak in grains in the YDB. However, such testing works only under the most favorable conditions, such as in loose, dry sediment with a high concentration of grains. If the samples contain high percentages of clay or are damp, we found that the magnet performs poorly. In addition, even if conditions are ideal but the concentration of grains is low, such as <1g/kg, we found it difficult to quantify the amount of grains on the magnet in the field. In summary, we found it far simpler to locate the YDB by analyzing magnetic grain abundances in the laboratory following the procedures below. We used only grade-42 or higher neodymium magnets, having found that nearly all other magnets are too weak and that some will completely fail to extract any magnetic grains. Typically, we used the size 2 ´1 ´ 0.5 inches (1 inch = 2.54 cm), which was convenient for field and laboratory work.

Although sonication is a common way to separate magnetic grains, the process was not used in our studies, because the procedure typically collects only the smallest and most magnetic grains, excluding up to 90% of the remainder, including many of the most interesting items, such as titanium-rich microspherules.

Typically, several methods were used to separate magnetic grains from sediment, depending upon the type of sediment. For large-scale processing, the following basic procedures were used with automated equipment and a bank of magnets, which were placed in a moving stream of either wet or dry sediment. Small samples were processed manually.

Loose or sandy sediment. About 500-1,000 g of friable sand or silt was first dehydrated at room temperature and weighed, and then, the samples were put into a container and the lumps were broken up. All of the processing was done using non-metallic tools to avoid adding foreign metal to the sample, and care was taken not to crush the fragile carbon component, if it was to be extracted also. Next, the magnet was placed in a 4-mil plastic bag to prevent grains from sticking directly to the magnet. A sediment sample was poured over the tightly bagged magnet into an empty container. Magnetic grains stuck to the magnet, and when the magnet was removed from the bag, the grains fell into a separate container. The process was repeated until nearly all of the grains were recovered.

Final step. As an essential final step to remove dust and debris, which can conceal the magnetic grains and spherules, the magnetic fraction was placed in a beaker of water. Then, the bagged magnet was gently agitated in the beaker to attract the magnetic grains. These were then deposited on a dry lab dish, by touching the wet bag to the plate after the magnet was removed from the bag. After drying, the sample was weighed, catalogued, and examined microscopically at ´100-150 magnification.

Sticky or clayey sediment. For sediment that was difficult to pulverize, we added ≈4 liters of water to each 500-1,000 g of sediment and homogenized it into slurry. The bagged magnet was then used to extract magnetic grains from the fluidized mixture. The magnetic grains were then released from the magnet into a separate container of water and then retrieved onto a laboratory dish as in the final step discussed above.

Extraction of magnetic microspherules. To find microspherules, the magnetic fraction was extracted from a weighed sediment sample with the neodymium magnet. We found it essential to complete the final step of cleaning the magnetic fraction with water, as outlined above. Also, because there are relatively few microspherules in bulk sediment, it was often necessary to inspect the most or all of the magnetic fraction extracted from 500-1,000 g of sediment. Next, one or more ≈100 mg aliquots of the magnetic fraction were weighed, dusted sparsely across a microscope slide, and scanned microscopically. Microspherules, which typically ranged from 10-100 mm, were counted, and abundances were extrapolated to quantity per kg. While viewed at ´100-150 magnification, selected microspherules were removed from the magnetic fraction manually with a moistened probe and placed onto an SEM mount or double-sided tape on a microscope slide. These spherules were either left whole or sectioned and given a microprobe polish for analysis by laser ablation or x-ray fluorescence (SEM/XRF).

Extraction of carbon spherules, glass-like carbon, and charcoal. Carbon spherules have a low specific gravity, and water floatation was used to assist with their separation. Typically, one kg of sample was added to ≈4 liters of water and agitated. The floating fraction was captured with a 150-mm sieve. In addition, there was often a carbon fraction with a specific gravity slightly higher than that of water, and that was removed from the top of the wet sediment visually. After drying at low temperatures, the carbon spherules were collected either visually or gravimetrically by vibrating the dried sample on an inclined, polished surface. Glass-like carbon and charcoal, contained in the same sample, were extracted manually and weighed.

Radiocarbon. AMS radiocarbon dating was performed by J. Southon (Keck Carbon Cycle AMS Facility) on peat and silt from Lake Hind. The radiocarbon date was converted to calibrated dates using IntCal04 (11).

Inductively coupled plasma mass spectrometry (ICP-MS) analysis. ICP-MS analyses. The isotopes evaluated for this investigation were: 52,53Cr ; 58,60,61,62,64Ni ; and 191,193Ir . Uncertainties varied by isotope, but all were less than ±20%. These isotopes were selected to evaluate the possibility of ET material in the sediment samples. Only Ir showed anomalous values. More details on the rationale for the selection of these isotopes, the ICP-MS conditions, analytical details of other isotopes not reported here, and the results and basis of the elements selected for further study will be presented in a forthcoming paper. This suite of isotopes allowed the use of aqua regia type acid mixtures to facilitate digestions. The digestion scheme allowed elements on the outside versus inside of the particles to be studied separately.

The analysis process involved digestion with concentrated Fisher OPTIMA nitric acid (HNO3) and then concentrated Fisher OPTIMA hydrofluoric acid (HF) with evaporation of the hydrofluoric acid before ICP-MS analysis in 5% (vol/vol) HNO3. All vessels and containers were acid washed in 10% nitric acid overnight, rinsed with ASTM I water, and dried beforehand.

Digestions. Initially, large sample weights of ≈100 g were used to screen the various isotope ratio changes to detect changes in uranium (U) isotopes. A method blank and a positive control (National Institutes of Standards and Technology, Buffalo River Sediment SRM 8704) were analyzed in parallel.

Screening digestions. Each 100-g sample ground in a mortar and pestle to pass through a 149-mm sieve was allowed to digest overnight in 75 ml of concentrated nitric acid in a Teflon beaker of known weight in a fume hood. The temperature on a hot plate was stepped-through for 2 h with Teflon watch glasses on at 50-55°C, 70-75°C, 80-85°C, 90-95°C, 100-105°C, and 110-115°C and then allowed to reflux with Teflon watch glasses until there were no more brown fumes. The gradual ramp was necessary to avoid boil-over and bumping of the heterogeneous digestion mixture. This took up to a week. After cooling, 75 ml of concentrated HF was added and allowed to stop bubbling. After 2 h, the above temperature ramp was repeated with watch glasses on after adding another 45 ml HF. The watch glasses were removed and the 2 h temperature steps then done at 125-130°C, 135-140°C, 145-150°C, 165-170° C, 175-180°C, and 195-200°C until dry. After cooling, the residue was weighed, and broken up with an acid washed pestle/Teflon spatula. The same process was repeated on the residue with 60 ml of concentrated HF alone, and then another 60 ml.

The dried broken up digestion residue (usually between 36% and 78% of the original weight) was extracted with 5% nitric acid in 2-h steps at 50°C, with each liquid being combined by decantation after cooling in a small 150-ml Teflon beaker where the combined dilute nitric acids extractions were evaporated at 110-115°C. This amalgamate was evaporated to ≈15 ml and used for analysis after cooling. This solution had a precipitate after standing and the liquid portion was carefully decanted into another centrifuge tube that was then centrifuged at 900 ´ g. The solids of the dilute HNO3 extraction steps and the centrifugation step were combined, dried to constant weight, and set aside for further analysis by nondestructive analytical chemistry techniques. This residual material was ≈27-64% of the original weight. The centrifuge tube supernatant was analyzed by ICP-MS, and its dry weight was ≈3.5-5.6% of the original weight. The solutions were yellow, orange, or red to orange compared to colorless for the method blank, and orange for the NIST sediment sample.

After the initial screening results were analyzed, small amounts of samples of 1 g were then digested to provide a HNO3/HCl available fraction, a HF available fraction after nitric acid digestion, and a residual fraction.

Small-weight digestions. Approximately 1 g of the sample ground in a mortar and pestle to pass through a 149-mm sieve was digested twice in a Teflon beaker with concentrated HNO3 (30 ml at room temperature for 16 h, 55-60°C for 16 h, 85-90°C for 16 h, then 135-140°C for 16 h before cooling and decanting the supernatant into another Teflon beaker, followed by 20 ml of concentrated nitric acid digestion for 16 h at 135-140°C) followed by three digestions with 4:1 HCl/HNO3 (20 ml ´ 3, each for 16 h at 135-140°C). Each residue with solids was dried at 135-140°C before the next digestion (SR1). The final solid residue after the last 4:1 HCl/HNO3 leaching was also retained. The 5 extractions were combined, evaporated, and the dry broken up residue leached with 30 ml 5% (vol/vol) nitric acid at 100°C to constitute the soluble phase that was centrifuged for 10 min at 900 ´ g. The solid residues (from leaching with 5% nitric acid and centrifugation) were combined with SR1 for the HF digestions

Two HF digestions followed, one with 4:1 HF/HNO3 (one with 30 ml for 16 h at room temperature, 55-60°C for 16 h, and then 135-140°C for 16 h) and decanting the supernatant into another Teflon beaker. The second digestion of the dried residue with particulates was with 30 ml of concentrated HF, and the second leaching was combined with the first extraction, dried, and the solid then leached with 5% nitric acid, and the leachate centrifuged. The centrifugation solids and leachate solids were combined with the solid from the second HF digestion step and then dried. The Teflon beakers were then scraped with a Teflon spatula to provide the residual solid weight that varied from 0.02% to 10% of the original weight. These residues were further analyzed by nondestructive instrumental analysis. The colors of these residual solids varied from black/greasy and black/hard to white flake (samples) to yellow-orange/cream (method blank and the National Institute of Standards and Technology sample). The digestion of small samples was thus at least 90% efficient in digesting the original sample.

X-ray Fluorescence (XRF). Representative microspherules were sliced, polished, and mounted for analysis by XRF with a scanning electron microscope (SEM) by B. Cannon (Cannon Microprobe). The x-ray spectra were obtained using an ARL SEMQ Electron Microprobe operated at 20 kV accelerating voltage, 50-nA beam current, 52.5° x-ray take off angle, a Kevex 2003 energy dispersive x-ray detector (EDS) biased at 620 V with an 8-mm-thick beryllium window and a PGT MCA 4000 multichannel analyzer. The resolution of the detector is 159 eV at Mn K alpha. Elements with atomic number 10 and smaller are not detected by this system due in part to thickness of the beryllium window on the detector. Different regions of the microspherules were randomly analyzed to obtain average elemental concentrations.

Prompt gamma-ray activation analysis (PGAA). PGAA of samples from many sites was performed at the Department of Nuclear Research, Institute of Isotopes in Hungary. PGAA is a non-destructive technique (12), using neutron beams to excite the samples producing gamma-ray spectra unique to each element. Typically, several gamma-rays are excited for each element, which can be used for analysis. PGAA is sensitive to the main constituents, except oxygen, and many trace elements in a sample. Concentrations are normalized to the total sample composition assuming standard oxidation states. Bulk samples of magnetic grains and microspherules, ranging in size from 9 mg to 13 g were analyzed with PGAA for H, B, F, Na, Al, S, Si, Mg, Cl, K, Ca, Sc, Ti, V, Cr, Mn, Fe, Co, Ni, Cu, Cd, Sm, Eu, and Gd. Uncertainties varied by element, but all were less than ±20%.

Nneutron activation analysis (NAA). The analysis of samples from many sites was performed at Becquerel and Activation Laboratories in Canada and at the Department of Nuclear Research, Institute of Isotopes, in Hungary. NAA was used to analyze trace element concentrations in both bulk sediment and magnetic grain samples, which were analyzed for Na, Si, Ca, Sc, Cr, Fe, Co, Zn, As, Se, Br, Rb, Zr, Mo, Ag, Cd, Sn, Sb, Te, Cs, Ba, Ce, La, Nd, Sm, Eu, Tb, Yb, Lu, Hf, Ta, W, Ir, Au, Hg, Th, and U. Uncertainties varied by element, but all were less than ±20%.

1. Haynes CV, Jr. (2007) in Murray Springs: A Clovis site with Multiple Activity Areas in the San Pedro Valley, Arizona, eds, Haynes CV, Jr, Huckell BB (Univ of Arizona Press, Tucson), pp 240-249.

2. Haynes CV (1995) Geoarchaeology 10(5):317-388.

3. Kooyman B, Newman ME, Cluney C, Lobb M, Tolman S, McNeil P, Hills LV (2001) Am Antiq 66:686-691.

4. Goodyear A (2005) In Paleoamerican Origins: Beyond Clovis, eds Bonnichsen R, Lepper BT, Stanford D, Waters MR (Texas A&M Univ. Press, College Station), pp 113-132.

5. Editor. (2000) Mammoth Trumpet 16(1):1-4.

6. Waters MR, Stafford TW, Jr (2007) Science 315:1122-1126.

7. Erlandson J, Kennett DJ, Ingram BL, Guthrie DA, Morris DP, Tveskov M, West GJ, Walker P (1996) Radiocarbon 38:355-373.

8. Boyd M, Running GL, Havholm K (2003) Geoarchaeology 18(6):583-607.

9. Boyce JI, Eyles N (1991) Geology 19:787-790.

10. Sharitz RR, Gibbons JW (1982) The Ecology of Southeastern Shrub Bogs (Pocosins) and Carolina Bays: A Community Profile (U.S. Fish and Wildlife Ser, Washington, DC), pp 93-94.

11. Stuiver M, Reimer PJ (1993) Radiocarbon 35:215-230.

12. Molnar GL, ed (2004) Handbook of Prompt Gamma Activation Analysis (Kluwer Academic, Boston, MA), pp 423.






North Carolina Antique Map Gallery


Its Spring and

that means



John Lawson's Trek Through the Carolinas 

In 1709 



The News & Observer


An Op-Ed I recently wrote



Archibald D. Murphy

And the improvement of North Carolina in 1815


Penguins and Parrots in North Carolina?


camp.gif (6644 bytes)