Response to Comments by John R. Southon, Lawrence Livermore National Laboratory, and R.E. Taylor, University of California, Riverside.

William Topping
Principal Investigator, NSF-Physics 9986999

A Mammoth Trumpet article (16(2): 9-16) entitled "Terrestrial Evidence for a Nuclear Catastrophe in Paleoindian Times" which is the subject of scrutiny by the authors above in the17(2) issue of the same journal resulted from 11 years of interdisciplinary research by the Principal Investigator. There is misinterpretation by authors Southon/Taylor who rebutted the piece in asserting they could find only 1 young date, and no related evidence such as extinctions and mutations. We (Firestone and Topping) presented 7 definite Paleo-Indian sites with "extremely young" radiocarbon dates: Gainey, Leavitt, Zander, Thedford, Potts, Alton and Grant Lake. As a factual matter there "is" a pattern to abnormally young dates in the Great Lakes region of North America. The sources for these dates were documented thoroughly, and clearly presented in the article. Among Paleo-Indian specialists (relatively few), this pattern is well-known. In addition, we discussed extinctions and mutations that are chronologically related, and based on evidence for prehistoric neutron flux most likely a direct result.

"Depleted" Uranium 235 is caused by neutrons, and the excessive Plutonium 239 found in various Paleo-Indian artifacts also is attributable to "neutrons in prehistory" considering all available evidence. The genesis of 14C is neutrons (14N + n = 14C), and the fact that we have hard evidence for prehistoric neutrons necessarily means 14C had to have been produced both in situ, and in the atmosphere. Many sediment profiles have been taken at widely separated locations in Michigan and elsewhere, and gamma analyzed. LBNL provided gamma runs accessed and analyzed with Maestro 5.30 software. These gamma spectra reveal anomalies at the B-C sediment interface, or the Paleo-Indian horizon at depth. Recent radiochemistry results on a Leavitt artifact "acid-etched" to remove the outer ~ 1/4 of the heavily patinated (silicate coating) surface, which eliminates the possibility of contamination from fallout (we tested this hypothesis with long-exposed Bayport chert: "no support" for fallout hypothesis), indicate Plutonium 239 at gross abundance. A massive prehistoric neutron event is indicated, and the micrometeorites themselves have been the subject of intense study which has included plutonium radiochemistry and gamma analysis on samples which were acid-bathed and washed with distilled water to remove all carbonates and adhering sediments. The micrometeorites exhibited the following attributes: the presence of both Plutonium 238 and 239, Cesium 137, and "depleted Uranium 235." Since the half lives of Plutonium 238 and Cesium 137 are so short (~ 87.7 and ~ 30.07 years respectively) the possibility of an exotic form of ongoing fission has been dismissed a priori in favor of the "fallout hypothesis" despite the logical inconsistency involved in attributing "depleted Uranium 235" to fallout, which is impossible.

Most persuasively, at Phoenix Memorial Laboratory (University of Michigan) a rigorously controlled gamma analysis was undertaken in 1996. The methodology compared 3 Paleo-Indian artifacts (Butler, Gainey and Leavitt retouch/channel flakes which are thin) to controls in the form of 2 later-period artifacts from the same geographic area (higher stratigraphic positions in sediments) and 4 separate "chert" controls from different sources representing respective chert types (Bayport and Upper Mercer), and then gamma runs for identical time periods to control half-lives ( t1/2 234Th = ~ 24 days, and t1/2 235U = ~ 7.04E8 years). We attempted to achieve nearly identical dimensions in terms of thickness, and configuration. Uranium is basically 238U and also 235U which is present in standard abundance at ~ 0.72%. The gamma "lines" or KeV energies which indicate uranium proportions are ~ 63, ~ 93, ~ 144, and ~ 186. The lines at 63/93 and 144/186 are key indicators since ~ 144 is basically 235U with an intensity (basically, how easily the gamma energy can be identified) of ~ 10.96 at 143.678 complicated somewhat by 223Ra at 144.432 with a lower intensity of ~ 3.22. The line at ~ 186 is mostly 235U at 185.712 with an intensity of ~ 57.2, again complicated somewhat by 226Ra at 186.221 with a much lower intensity of ~ 3.59. The lines at ~ 63 and ~ 93 are essentially all 234Th, the daughter of 238U, with the respective associated intensities of ~ 4.8 at 63.29, ~ 2.891 at 92.38, and ~ 2.77 at 92.80 (1).

Comparing the proportions (nets) is very revealing since "depleted 235U" would be most manifest in the line at 186 compared to the lines at 63 and 93. In Phoenix runs the 144 line was not discernable, but results are consistent and probably conclusive. Uranium content in cherts varies; Bayport is relatively rich, while Upper Mercer is not. Butler is ~ 1 km away from Gainey, while Leavitt is ~ 80 km from Gainey. At Gainey, Upper Mercer was the chert type while at Butler and Leavitt the chert type was Bayport. The other two artifacts (Archaic and Woodland) were made of entirely different chert types.

TABLE 1.  Phoenix Memorial Laboratory gamma runs on specimens.
129,600 seconds each run.  Paleo-Indian in RED.  Net counts, background out.
                        234Th      234Th    mostly 235U               % 235U
Sample      Wt/mg        63         93          186       Total      186%/total      Conclusion 
Gainey      127.5       1645       2655        1121        5421      20.6788415      depleted 235U
Mercer 1    713.8       1694       2853        1365        5912      23.0886333      normal 235U
Mercer 2    798.5       1876       2763        1532        6171      24.8257981      normal 235U
Butler      899.8       2130       2864        1156        6150      18.796748       depleted 235U
Bayport 1   193.4       1948       2752        1441        6141      23.4652337      normal 235U
Bayport 2   236.4       1930       3060        1516        6506      23.3015678      normal 235U
Leavitt     114.8       1931       2509        1112        5552      20.0288184      depleted 235U
Lakewood    126.3       1713       2742        1468        5923      24.784737       normal 235U
Pinegrove   246.2       1784       2795        1478        6057      24.401519       normal 235U

The results require little "analysis" since the proportions obviously are so different, and very unlikely (essentially impossible) due to coincidence. Other LBNL runs were analyzed, which did pick up the 144 line, and also indicated "depleted 235U," consistent with McMaster Nuclear Reactor and Phoenix results. Compared to a uranium ore sample (186 at ~ 55.30804% of 63, 93, 144 and 186), the 186 line in a Butler artifact came in at ~ 19.82593% (essentially identical to Phoenix results), and Taylor artifacts at ~ 29.93631% which is precisely the pattern expected by latitude/longitude. Gainey "C" sediments came in at ~ 34.7042% for the 186 line, and the micrometeorites came in at ~ 31.5792%. There is one more indicator of internal consistency in these data. Specifically, the 235U depletion rates for Gainey and Butler indicate more depletion in Butler. These two Paleo-Indian sites are very close geographically, but the Butler artifact was made of Bayport chert with more uranium than Upper Mercer (Gainey). In a neutron event, higher uranium content would lead to more nuclear interactions (i.e., more secondary neutrons produced) which would lead to higher rates of 235U depletion which is exactly what the Phoenix results indicate.

These results are especially meaningful since "artifacts" and "sites" should be considered in their larger context as entities representative of the overall environment during a "particular window in time," but only at a certain longitude/latitude/depth, just as the K-T boundary is anywhere in the world. The Phoenix, LBNL, McMaster and other results are critically important since they do not record data about "particular oddities" but rather provide coherent and general information about the prehistoric North American and world landscape just as the ice cores at particular depths do.

One test did conflict. A TIMS (thermal ionization mass spectrometry) analysis at USGS, Menlo Park, found "no depletion" while McMaster results (delayed neutrons) found depleted 235U, completely consistent with gamma comparisons and outsourced radiochemistry. In a test of the TIMS method itself, "depleted uranium" in U3O8 powdered form from Starmet Corporation was obtained. A sample was sent to a TIMS laboratory, and the results came back as ~0.23% 235U. In fact, the depleted uranium was supposed to be ~ 0.20%, or less. In two LBNL gamma runs, results on the same depleted uranium came back as ~ 0.20% 235U, or precisely that expected. In regard to the Menlo Park TIMS analysis, 6 samples were analyzed and in the case of Onondaga chert (USGS lab. no. 01-141, July 9, 2001) the result was "no U yield" (2) which is difficult to explain since neutron activation serial analyses detected uranium in Onondaga chert in respective amounts of 0.436 ppm (3) and 0.463 (4) and the standard deviations for the series were 0.100 and 0.109, again respectively. Formerly, TIMS was regarded as more accurate than neutron activation analysis which detects on the basis of fission/capture byproducts (and delayed neutrons for 235U determination in the McMaster method). There now are larger questions involved regarding respective methodologies, and the Menlo Park results in particular must be regarded as meaningless since they are impossible to reconcile with the available NAA results on Onondaga chert, but unfortunately arrived before the most recent data became available.

In the case of McMaster, of "thousands of runs" only the Paleo-Indian artifacts and related chert samples evidenced depleted 235U. Importantly, a uranium standard in the form of uraninite dissolved in HF and then reincorporated into a granular absorbent with a total gamma activity of ~ 100 Bq was analyzed at McMaster. This standard was supplied in spherical form, and McMaster quartered the sample and in 4 separate analyses (each ~ 25 Bq which is near "background" in some glaciated sediments tested) the 235U proportion came in at ~ 0.72% in each case. If McMaster methodology was flawed, then we should not have found the standard 235U abundance in these 4 control tests of the method. Again, it effectively is statistically impossible for all of these results to be "coincidence," and the bulk of the evidence in the form of McMaster methodology, gamma analyses, and associated radiochemistry are consistent and coherent in detecting a prehistoric neutron event. In this regard, the Principal Investigator owns an NaI (sodium iodide) detector and since 1998 has done many (~ 10,000) gamma runs with a variety of uranium ores, thorium, and sediments/rocks/artifacts which has provided much expertise.

Since the inception of the investigation, the Principal Investigator has regarded a "solar flare" as direct cause for the clear pattern at ~ 12,500 yrs bp as recorded in ice and marine data, Paleo-Indian radiocarbon dates/artifacts and associated extinctions and mutations, and also sediments at depth simply because the overall pattern suggests an "event" of less than or about 24 hours duration with primary observable effects consistent with the rotation of the earth in respect to the sun. Earlier "cosmic events" appear to have contributed to isotopic anomalies. There is an inverse relationship between micrometeorite densities in artifacts and related Paleo-Indian radiocarbon dates. At Gainey and Leavitt where the radiocarbon dates obtained were virtually identical at ~ 2880 bp rcy the particle density was ~ 70,000 cm^2. Particle densities at other sites are lower by latitude/longitude, with older associated radiocarbon dates. At Taylor, for instance, the density of micrometeorites was ~ 40,000 cm^2. This implies particles were deposited in one shockwave of great intensity, with necessary associated neutrons. The most likely culprit is the sun, but this may be incorrect. Independent testing to confirm depleted 235U in the micrometeorites had been considered to "rule out" a supernova since the debris ought to contain "enriched" 235U (assuming uranium actually is formed in supernovae), but now has been discarded because the debris itself would have been subjected to neutrons, and hence depleted "then." Firestone may be correct in hypothesizing a supernova, and there are other possibilities as well.

The characteristics of a shockwave with associated nuclear interactions are chaotic, but the Southon/Taylor critique of the 14C production involved two fundamental oversights. First, the assumption that the Pleistocene atmosphere contained ~78% nitrogen is entirely problematic. With ice having covered almost half the earth for a long period of time (perhaps approaching ~ 1.6 million years), the atmosphere must have been less dense due to lowered biotic activity, and nitrogen content presumably was substantially lower. Second, in a shockwave the relative densities of atmospheric compression into in a "funnel" would have involved entirely different nuclear interactions by latitude/longitude/height including production of nitrogen oxides, and the most probable (essentially, "definite" because of the 14C + n-t,f reactions) associated destruction of 14C, which itself decays back into 14N (t1/2 = ~ 5730 years). Literature on atmospheric nuclear weapons testing makes it clear (5; see 1.36, 2.131, 2.132, 2.140 and 2.150 in particular). Neutron/gamma response to varying degrees of atmospheric density in a shockwave are complicated and symbiotic, and in all fairness to everyone probably "never" calculable in mathematical terms since the event (main) is deep in time, and environmental conditions then so different from modern.

Decisively, at the Lewisville, Texas, Paleo-Indian site there is in fact a perfectly reliable radiocarbon date of ~ 26,600 bp rcy obtained in 1985 (6). It is common knowledge in the archaeological community that Paleo-Indian artifacts are distinctive, and cultural markers of a "window in time." Since the Lewisville site is "at least" 26,000 years old, younger Paleo-Indian radiocarbon dates cannot possibly be correct, and the construct that maintains Paleo-Indian is confined to ~ 12,000 years therefore also must be incorrect, and should be discarded immediately. Using common sense, if prehistoric peoples arrived in North America first and then migrated to Central and South America, 14C dates at lower latitudes "must be" younger. In fact, South American radiocarbon dates approach ~ 30,000 bp rcy. Lewisville "fits" while younger Paleo-Indian dates in North America do not, and therefore all must be incorrect (for some reason).

Radiocarbon dating works "perfectly well." However, it derives dates from the amount of 14C present. But, there definitely are anomalies by latitude/longitude/depth in sediments that authoritative archaeologists have noted for many years. It is clear that both communities are correct, and interdisciplinary collaboration in the pursuit of scientific truth stongly is encouraged. Additionally, this short piece hopefully will serve as a reminder to all, both avocational and otherwise, to be more "open-minded" in the pursuit of patterns to factual scientific data rather than adherence to "dogma," and for everyone to consider data within the broader framework of a holistic (rather than "particularist/reductionist") point of view.


1. Firestone, Richard B. Table of Isotopes, Eighth Edition, 1998 Update (CD ROM, Coral M. Baglin and S.Y. Frank Chu, eds.).

2. Bischoff, James L. 238U/235U ratio measured on chert samples by thermal ionization mass spectrometry. USGS, Menlo Park, 9 July, 2001 (ms. possession of author).

3. Luedtke, Barbara E. An Archaeologist's Guide to Chert and Flint. Archaeological Research Tools 7, Institute of Archaeology, University of California, Los Angeles, 1992.

4. Jarvis, H. INAA characterization of of Onondaga chert; a preliminary study; in western New York. Masters thesis, Department of Anthropology, State University of New York, Buffalo, 1988, referenced in Luedtke.

5. Glasstone, Samuel and Dolan, Philip J. (eds.), The Effects of Nuclear Weapons, Third Edition. United States Department of Defense and Energy Research and Development Administration; 1977.

6. Shiley, Richard H., Randall E. Hughes, Richard A. Cahill, and Kenneth L. Konopka.
Moessbauer Anaysis of Lewisville, Texas, Archaeological Site Lignite and Hearth Samples. Illinois State Geological Survey, Environmental Geology Notes 109. Carbondale, Illinois; 1985.