Dr. Henke emphasizes the assumed 4.5 billion year age of the earth, which is not my main concern. I am mainly concerned with the age of life on earth.
Dr. Henke gives some technical corrections to my article, which I appreciate, but these do not affect the main argument.
Dr. Henke gives evidence that the decay rates have not changed.
He says,
I ordered Dalrymple's book (Dalrymple, G. B., 1991 The Age of the Earth Stanford University Press, 1991), but unfortunately it says nothing about the effect of neutrinos on decay rates. The agreement of many different techniques on meteorites suggests that any change in decay rates would have to affect all of them proportionally, which seems unlikely for changes based on large numbers of neutrinos. However, there is reason to believe that a change in the speed of light might also change decay rates, and cosmologies permitting a larger speed of light in the past have been proposed by some scientists.Dalrymple (1984, p. 88-89) refutes in some detail the various creationist claims that radioactive decay rates may be influenced by neutrinos, neutrons, and cosmic radiation, including Dudley's "neutrino sea."
Dalrymple (1984, pp. 88-89) discusses various creationist proposals for altering decay rates involving neutrons or neutrinos. He rejects the idea that neutrinos could affect radioactive decay, but really gives no support for this statement. He rejects Dudley's hypothesis of a "neutrino sea," but this does not affect the possibility that neutrinos from some other source (perhaps a supernova) could influence radioactive decay. Concerning the identity of the supernova, Slusher (Slusher, H.S., 1981. Critique of Radiometric Dating, Institute for Creation Research, Technical monograph 2 (2nd ed.), 46 pp, p. 55) cites F.B. Jueneman (Industrial Research, Sept., 1972, p. 15) in the following speculation:
The online Encyclopedia Britannica has this to say about the Gum Nebula:The remnant of that local big bang is a pulsar called Vela-X (PSR 0833-45), which recent observations have positioned in the southern sky some 1,500 light years away, and which is considered to have given rise to the huge Gum Nebula ... Being so close, the anisotropic neutrino flux of the super-explosion must have had the peculiar characteristic of resetting all our atomic clocks.
Concerning the timing of this supernova, if the Vela Pulsar is the same as Vela-X, then the estimated date of the supernova would be fairly recent, according to the online Encyclopedia Britannica:Gum Nebula, largest known nebula in terms of angular diameter as seen from Earth, extending over at least 40 degrees in the southern constellations Puppis and Vela. A complex of diffuse, glowing gas too faint to be seen with the unaided eye, it was discovered by the Australian-born astrophysicist Colin S. Gum, who published his findings in 1955. The Gum Nebula lies roughly 1,000 light-years from the Earth and may be the remnant of an ancient supernova--i.e., violently exploding star.
The Crab Pulsar is the youngest known, followed by the Vela Pulsar that has a projected timing age of 11,000 years.
Concerning a possible change in decay rates, Dalrymple (1991, p. 329) mentions that Pb-Pb ages for meteorites tend to be about one percent higher than Rb-Sr and K-Ar ages. This can be explained by the K-Ar ages measuring a slightly later event, or by uncertainties in the Rb-Sr decay constant. It can also be evidence of a change of the decay constants, since Pb-Pb ages are based on the decay of uranium to lead, which partially involves alpha decay, and Rb-Sr decay is based on beta decay (ejection of an electron). K-Ar decay is based on electron capture, which is in a sense the inverse of beta decay. So it could be that a change in decay constants would treat K-Ar decay and Rb-Sr decay the same, but U-Pb decay would be affected differently. In fact, Slusher (1981, p. 38) states "in order to match K-Ar ages, which average somewhat higher than the U-Th-Pb ages, to the latter ages, the value 0.08 is arbitrarily taken. However, this does not remedy the situation, and the ages are still too high." This indicates that there may be a systematic difference between U-Pb ages and K-Ar ages, which could be a further evidence of a change in decay rates. The Pb-Pb method essentially depends on the ratio between decay rates of two uranium-based decay systems, both involving alpha decay. Thus the Pb-Pb method may be less sensitive to changes in the decay constants than the U-Pb methods, and perhaps this is one reason why it yeilds better agreement with other methods.
A difference of one percent in an age of 4.5 billion years may not seem like much, but it amounts to about 45 million years. If a catastrophic event increased Pb-Pb ages by 600 million years and K-Ar and Rb-Sr ages by 550 million years, then this would produce the observed one percent discrepancy in meteorites, but also invalidate ages on Cambrian and later strata.
It is also interesting in this regard that beta decay creates an anti-neutrino. It could just as well occur by absorbing a neutrino, and thus an increase in the flux of neutrinos could increase the beta decay constants. The web site How to Change Nuclear Decay Rates mentions that electron capture invoves the emission of a neutrino, and so it might be stimulated by the absorption of an anti-neutrino. In fact, since neutrinos are so difficult to detect, it could be that they are already causing some of the radioactive decay that is observed.
Slusher (1981, p. 23) cites (Lee, T.D. and Yang, C.N., Phys. Rev. Lett. 4, 1960, p. 307) as predicting that "nuclear reactions may be induced by electron neutrinos." Slusher (1981, p. 23) mentions that the transformation of 37Cl to 37Ar by a neutrino "is an established method of study of high-energy solar neutrinos." This transformation involves the change of a neutron to a proton. The same reaction would cause the decay of rubidium to strontium. In fact, this reaction for rubidium might be much more likely to be caused by neutrinos than the transformation of 37Cl to 37Ar, since rubidium naturally decays to strontium, so the energy needed for the decay of rubidium is smaller than for 37Cl. Thus Rb-Sr ages might be increased significantly by large numbers of neutrinos. The fact that such reactions can be much more likely for some nucleii than others is shown by the following statement from Dalrymple (1984, p. 94) concerning reactions with neutrons: "the neutron-capture cross sections for 206Pb and 207Pb are not equal, as Cook ... assumes, but differ by a factor of 24 ... ." Thus it is likely that Rb nucleii might capture neutrinos and decay much more readily than 37Cl.
The same reaction would cause the decay of carbon 14, and might increase radiocarbon ages as well. However, since carbon 14 has such a short half life, and decays spontaneously relatively rapidly, the effect due to neutrinos may be smaller than for substances such as rubidium with very long half lives. This might explain why low carbon 14 ages are sometimes found in conjunction with lava flows having very old radiometric ages. By analogy with the effect of neutrinos on decay, anti-neutrinos could change a proton to a neutron, and thus contribute to the decay of potassium to argon, increasing K-Ar ages. Neutrinos, through the weak interaction, may be able to disturb nucleii slightly, and thus induce unstable nucleii to decay even by alpha decay. Neutrinos interact with matter through W+, W-, and Z bosons, which carry the weak force, and perhaps through an exchange of these particles with a proton or a neutron, a neutrino could perturb a nucleus without the identity of either particle being changed. Maybe the reason that too few solar neutrinos are observed is that most of them are absorbed in the sun and contribute to further radioactive decay there. Another possibility is that a change in the speed of light could influence decay constants. (See http://ldolphin.org/barrychron.html ).
Concerning the possibility that neutrinos can induce nuclear reactions, we have the following quotation from the online Encyclopedia Britannica article "atom: nuclear reactions:"
Thus it appears possible that neutrinos, through inelastic scattering, could even induce alpha decay.Nuclear reactions also can be induced by leptons such as electrons and neutrinos. The probability for an electron-induced reaction is smaller than for a hadronic reaction because of the difference in forces. The main process is inelastic scattering, in which the nucleus is excited to a higher energy level. If the inelastic scattering transfers a large amount of energy to the nucleus (e.g., 20 MeV or larger), the nucleus becomes disrupted and ejects particles.
The weak interaction also causes reactions, but these are very difficult to observe. One part of the weak interaction changes a neutron into a proton or vice versa and brings about a corresponding conversion between an electron and a neutrino. The direction of the change is such as to conserve the total charge. An example of a neutrino-induced reaction is the reaction by which neutrinos were detected from a supernova in 1987. In this case, the neutrinos--or more precisely antineutrinos nu, --from the cosmic object interacted with protons of hydrogen atoms in a large tank of water. Positrons, e+, were created according to the reaction
nu bar + p ----> n + e+ The positrons were detected by the light they emitted as they traveled through the water.
One source said that supernovae emit 10^57 neutrinos, raising the possibility that neutrinos from a nearby supernova could influence decay rates. Other radiation (such as gamma rays) from a supernova might also influence decay rates.
If the neutrino flux contributes to radioactive decay, then we would expect to see variations in the rate of decay, corresponding to changes in the neutrino flux. This would imply that there would be statistical irregularities in radioactive decay. Such irregularities were claimed, according to Slusher (1981, p. 49):
The source for this information is Anderson, J.L., Abstract of Papers, 161st National Meeting, American Chemical Society, Los Angeles, 1971. In a similar vein, Slusher (1981, p. 26) reports:In a recent publication it has been reported that decay rates of 14C may not be predictable as previously believed ... . Dr. John L. Anderson performed experiments on molecular mono-layers with 14C added. The emitted radiations did not occur in the patten that classical theory assumes. Dr. Anderson's findings were confirmed through independent research at Atomic Energy Commission Laboratories.
Such irregularities were observed for carbon 14, cobalt 60, and cesium 137. The source for this information is Anderson, J.L. and Spangler, G.W., "Radiometric Dating: Is the `Decay Constant' Constant?", Pensee, p. 31. Even Dalrymple (1984, p. 88) recognizes such irregularities:Anderson and Spangler maintain that their several observations of statistically significant deviations from the (random) expectation strongly suggests that an unreliability factor must be incorporated into age-dating calculations.
Dalrymple cites the references Anderson, J. L., 1972, Non-Poisson distributions observed during counting of certain carbon-14-labeled organic (sub) monolayers, Phys. Chem. J. 76: 3603-3612 and Anderson, J.L.and G.W. Spangler, 1973, Serial statistics: Is radioactive decay random? Phys. Chem. J. 77: 3114 - 3121. Such statistical irregularities could be evidence that the neutrino flux is influencing the rate of radioactive decay.Under certain environmental conditions, the decay characteristics of 14C, 60Co, and 137Ce, all of which decay by beta emission, do deviate slightly from the ideal random distribution predicted by current theory ... , but changes in the decay constants have not been detected.
Dr. Henke criticized some statements in my article taken from Slusher
about the branching ratio for potassium. Slusher asserted that the
best known value of the branching ratio was not always used in
computing K-Ar radiometric ages. Unfortunately, Dalrymple (1991) says
nothing about the calculation of the branching ratio. He simply gives
the correct value for the K-Ar system. The issue is not just how well
this was known in the past, but which value was actually used, and
whether dates published in the past have been computed with the most
recent value. Often values for constants are standardized, so that
the values actually used may not be the most accurate known. All that
Dalrymple (1991) says is that his ages were all recomputed using the
most accurate values of the constants. This implies that some of them
were originally computed using less accurate values, which is similar
to Slusher's point.
Dalrymple (1984, p. 91) does discusses the branching ratio. He admits that Slusher's statements about it would have been true in the 1940's and early 1950's, but are no longer true. But he didn't say when the correct value for the branching ratio began to be used. Even some figures from (Faure, Principles of Isotope Geology, 1977) are based on another constant that is 2 or 3 percent too low, according to Dalrymple, and so there may be many ages in the literature that need revision by small amounts. However, Harland et al (1990) imply that nearly the correct value for the branching ratio has been known and used since the mid-fifties.
Dr. Henke criticizes my concern that argon can move in and out of
minerals:
However, Henke admits that this can happen in some cases. He states that geologists are aware of this problem, and make allowances for it. But it is more difficult to remove argon that has deposited on cracks in the mineral, which can be difficult to see. Dr. Henke referenced Davis A. Young frequently, but I was not able to find Young referenced in any of the other sources I examined except Dalrymple (1991). Dr. Henke states that hornblendes retain argon very well, but then later says that they can easily absorb excess argon.Dr. Plaisted wants to give his readers the impression that argon can readily move in and out of minerals and, therefore, the gas is too volatile for radiometric dating. Specifically, he quotes one of his anonymous friends that claims that argon easily diffuses from minerals (p. 11, also the identical statement is made in Slusher, 1981, p. 39). Of course, these statements are inaccurate generalizations. Young (1982, p. 101) notes that argon would more likely adsorb onto the surfaces of the minerals rather than move into their tight structures. Geochronologists are aware that excess argon may accumulate on mineral surfaces and the surface argon would be removed before analysis.
Geologists also recognize that heating causes argon to leave minerals, and that dissolved argon in a mineral that does not escape will become incorporated into it, artificially increasing its K-Ar age. I will comment more on this below, but a few comments now are appropriate. From a discussion of radiometric dating from Cornell, we learn (quoting approximately)
From another table at the same location, we find that argon loss in biotite by radial diffusion in an infinite cylindrical crystal of radius 150 microns, is 1/1000 of a percent in one year at 240 degrees centigrade, the same percent in 100 years at 200 degrees centigrade, the same amount in 10,000 years at 160 degrees centigrade, and the same amount in 10 million years at about 110 degrees centigrade. To lose one percent in one year requires a temperature of nearly 500 degrees centigrade.For a temperature of 300K (27 degrees C), there is no significant argon loss from biotite. At 600K (327 degrees C), there is a slow but significant diffusion rate. At 700K (427 degrees C), loss of argon is quite rapid.
Thus the temperature does not have to be very high for argon to move through rock. This also justifies Slusher's statements about argon moving in and out of rocks with ease. However, it does not seem likely that sedimentary rocks would be this hot very often, except near lava or magma flows.
But argon does not need to move through all rock in order to influence radiometric dates, it only has to reach ancient lava flows. This it can do by following the path of the ancient lava flow itself, coming up along the path of the magma. As the magma or lava cools, this path will consist entirely of hot magma or lava, and so the argon will have a free path, and will continue to enter the magma as it cools. Thus in many cases, the lava or magma will never completely degas, and extra argon will end up trapped in the cooled rock. This will result in artificially increased K-Ar ages.
Many ancient lava flows are relatively flat, in contrast to modern ones. Also, they appear to have been covered over quickly. The flatness means that the lava is a contiguous mass, and can still be reached from the hot magma by a continuous path of hot rock. The fact that they soon are covered over means that the argon has a hard time escaping vertically from the lava, so argon coming up from the mantle will tend to enter the cooling rock. Both facts will tend to produce artificially high K-Ar ages in these flows which will not be seen in modern lava flows in the same manner.
Modern lava flows often come down the sides of volcanoes, and thus become separated from their source by large distances. Also, they do not get quickly buried by additional sediment. Thus modern lava flows are not subject to the same mechanism of artificial increases in their K-Ar ages as are ancient ones. Also, it is reasonable to assume that as argon leaves the mantle in successive eruptions, the amount of argon remaining is reduced, so that later lava flows are less susceptible to such artificial increases in age. The path of magma also becomes longer for later flows, and the magma probably also is a little cooler, inhibiting argon flow. Thus later lava flows give younger K-Ar ages.
Another point to note is that even after it cools, the lava or magma may still have many cracks in it, permitting argon to flow. This argon will tend to deposit on the surface of minerals, but with the passage of time it will tend to diffuse into the interior, even if only a very small distance. This is especially true as the lava is cooling. This will make it more difficult to detect this added argon by the spectrum test described below.
Also, the diffusion of argon in cracks and channels of a mineral is likely much less temperature-dependent than diffusion through unbroken regions of the mineral, since diffusion through cracks and channels simply involves jumps through the air. By a combination of diffusion through cracks and channels, and short passages through unbroken regions of the mineral, argon may be able to reach a considerable distance into the mineral. At low temperatures, this may become the dominant means by which argon diffuses into a mineral, but the effect of this kind of diffusion at low temperatures may not be evident until many years have passed. Thus it may take experiments lasting 50 or 100 years at low temperatures to detect the effects of this kind of diffusion of argon, which however could be significantly increasing the K-Ar ages of minerals over long time periods.
Dickin (Radiogenic Isotope Geology, 1995, p. 247) mentions a study showing that volcanic rocks contain excess atmospheric argon, some of which cannot be removed by baking in a vacuum. Faure (1986, p. 74) also states concerning removing atmospheric argon before dating, "The difficulty can be reduced, if not completely avoided, by the removal of adsorbed atmospheric argon before the argon is extracted from the samples. It has been claimed that this can be accomplished by preheating samples under vacuum or by leaching them briefly with hydroflouric acid, or both ... . However Armstrong (1978) has questioned whether atmospheric argon, that has been acquired by minerals over a long interval of time, can be removed by this method."
Thus there is some means by which argon from outside can become very firmly embedded within a rock, and one would expect that the quantity of this argon would continue to increase over time, giving anomalously old K-Ar ages. Added atmospheric argon can be detected, because the ratio of argon 40 to argon 36 for atmospheric argon is 295.5 to one (Faure, 1986, p. 95), and the amount of argon 36 can be measured. But argon 40 coming up from the mantle and diffusing into a mineral would not be detectable in this way, because it has a higher ratio of argon 40 to argon 36. Faure (1986, p. 79) mentions that the ratio of argon 40 to argon 36 in the mantle may be as high as 10,000 or even 25,000 to one. Dickin (1995, p. 247) mentions that atmospheric argon adsorbed onto a rock can be as high as 70 percent of the total argon, especially for young material. This shows that rocks can adsorb a large amount of argon relative to the argon needed to give them old K-Ar ages, and also suggests that old K-Ar ages can be produced by external argon from the mantle. Over a long period of time, adsorbed argon will tend to diffuse into the rock, and thus it will be possible for even more argon to be deposited on the surface, increasing K-Ar ages even more.
Concerning excess argon, Faure (1986, p. 72) states:
Generally, excess 40Ar is observed in minerals that have been exposed to a high partial pressure of argon during regional metamorphism, in pegmatites ..., or kimberlite pipes. The argon that may either diffuse into the minerals or may be occluded within them is derived by outgassing of K-bearing minerals in the crust and mantle of the Earth. ... The presence of excess 40Ar increases K-Ar dates and may lead to overestimates of the ages of minerals dated by this method.
Concerning the abilities of various minerals to retain argon, Dr.
Henke writes:
However, there is a methodological problem connected with the manner in which geologists infer the argon-retention abilities of different minerals. Concerning the suitability of different minerals for K-Ar dating, Faure (1986, p. 72) writes "The minerals beryl, cordierite, pyroxene, and tourmaline frequently contain excess 40Ar, while hornblende, feldspar, phlogopite, biotite, and sodalite contain such excess 40Ar only rarely ... ." And how is this known? By comparing the K-Ar dates yielded by such minerals with the expected ones. Thus the correctness of the geologic time scale is assumed in deciding which minerals are suitable for dating. For example, concerning the use of glauconies for K-Ar dating, Faure (1986, p. 78) writes, "The results have been confusing because only the most highly evolved glauconies have yielded dates that are compatible with the biostrategraphic ages of their host rocks whereas many others have yielded lower dates. Therefore, K-Ar dates of 'glauconite' have often been regarded as minimum dates that underestimate the depositional age of their host." Dickin (1995, p. 56) says "in some cases erroneous glauconite model ages could be increased to near the stratigraphic age by leaching with ammonium acetate, which is thought to remove excess loosely bound Rb from the expandable layers of the lattice." But leaching with acetic acid had unpredictable effects. So for K-Ar dating, "the most highly evolved" glauconies are selected to give the best ages, and Rb-Sr model ages of glauconies can in some cases be improved by leaching with ammonium acetate, but not acetic acid. Model ages are based on assumptions about initial Rb and Sr concentrations, as well. Dickin (1995, p. 111) mentions how the processing of zircons in a magnetic separator is performed to obtain "spectacular increases in concordance." All of these choices are made in order to obtain dates that are more in agreement with each other. The result is that radiometric dating is in danger of being based on circular reasoning.Once Ar40 is produced by the decay of K40 within a mineral, the ability of the mineral to retain the argon would depend on the identity of the mineral and its thermal environment (e.g., Hyndman, 1985, p. 675-676). Hornblendes, for example, retain argon very well, while biotites are less effective and feldspars readily leak the gas. Geologists are familiar with the argon-retention abilities of different minerals and use these differences to their advantage.
Dr. Henke criticizes my concerns about the effects of heating and water
on K-Ar dating.
Here he confuses heating with metamorphism. The latter involves a change in the structure of a mineral, but lesser heating may not. However, heating can cause argon to leave a mineral, and in fact this is the basis for K-Ar dating in the first place. Faure (1986, p. 123) verifies that modest increases in temperature of one or two hundred degrees centigrade can have drastic effects on radiometric dates without any observable changes in the rocks. Faure (1986, p. 69) mentions specifically that heating can cause argon loss without any other physical or chemical changes in the rock. Dr. Henke also discusses the effect of hot water on rocks, producing alterations in their structure. Of course, a greater problem is diffusion of water through a sample that is not hot enough to alter the structure, and Dr. Henke does not discuss this. A further problem is that water can cause the transport of dissolved argon, which can then enter other rocks. He refers to "weathering," which can only occur for exposed rocks. Another problem is water traveling underground, which can also influence the argon content of rocks without weathering, as it is commonly understood.In several places in his report (for example, p. 6-7), Dr. Plaisted is concerned that radiometric dates may be unknowingly affected by the movement of water through the rock or by metamorphic heating. Fortunately, his concerns are largely unfounded.
Dr. Henke discusses xenoliths and xenocrysts, which are older rocks that can be carried along with magma. He says that geologists attempt to avoid these, but he does not say whether this is always possible, and whether some xenoliths and xenocrysts can be too small to see, even with a microscope. Dickin (1995, p. 249) mentions that some subaerially erupted lavas contain inherited argon: "Lavas shown by 14C dating of wood inclusions to be less than 1 kyr old nevertheless gave K-Ar ages up to 465 kyr." A cause that has been suggested for this is "partially digested crustal xenocrysts" (p. 250).
Dr. Henke discusses my concerns that various methods do not agree on the phanerozoic. This issue has already been discussed in our earlier exchanges. Let me just repeat here than many authorities point to common disagreements between various methods.
Concerning such disagreements, Dr. Henke states:
Along this line, it is interesting that Woodmorappe (Woodmorappe, J., Radiometric Geochronology Reappraised, Creation Research Society Quarterly, vol. 16, September, 1979, p. 102f.) mentions that there are a number of agreements between Rb/Sr and K/Ar dates that are considered "wrong" by geologists, who speculate that some other factor than a true age may be making these dates agree with one another.Young (1977, p. 190f) provides an excellent example of fossil data confirming the results of Rb/Sr and K/Ar dates for the Beemerville Nepheline Syenite in New Jersey. The syenite intruded into the Ordovician Martinsburg Formation, but it does not intrude into the overlying Lower Silurian Tuscarora Formation. Cross-cutting relationships, fossil data, and the geologic time scale indicate that the syenite should be 425 to 450 million years old. Rb/Sr and K/Ar dating of the syenite yielded dates of 424 ±20 million years, 436 ±41 million years, and 437 ±22 million years, which are reasonably consistent.
In his first reply, Henke states concerning disagreements between methods:
Unfortunately, Dalrymple (1991) has little or nothing to say about the phanerozoic. It is also possible that Faul (Faul, H., 1966 Ages of Rocks, Planets, and Stars, McGraw- Hill, New York.) is not mainly concerned about the phanerozoic.I cited Harland et al. (1989), Faul (1966) and Dalrymple (1991), which contain MORE than just "a few examples."
Dr. Henke further states:
This quotation seems to miss the point, as I did not question whether the different results were listed separately or averaged. A more important issue is the degree to which different methods agree.Dr. Plaisted (p. 15) recognizes that a majority of radiometric dates are K/Ar. He then expresses concern that an overreliance on the K/Ar method may lead to inaccurate dating of samples. ...
The K/Ar results may be averaged, but the literature overwhelmingly indicates that results from Rb/Sr, U/Pb and other methods would be listed separately, no matter how they compare with the K/Ar results (for example, see the table in Dalrymple, 1991, p. 140-141).
Dr. Henke acknowledges that discordant dates exist. A further problem with anomalous dates is that some dates may not be published if the articles mentioning them are rejected. I don't know about geology, but in many disciplines, publication is a difficult enterprise. Thus there could be more anomalies than we know about. Some work may not be published simply due to the time and effort it requires.
Dr. Henke discusses the question of atmospheric contamination of samples, altering their K-Ar dates:
My point in mentioning this is to show how easily argon can enter rocks. I realize that atmospheric argon can generally be corrected for. I am more concerned about contamination from argon that comes up from within the earth; this argon is almost entirely argon 40, and so it is nearly impossible to distinguish from radiogenic argon.Dr. Plaisted (p. 6-7 and elsewhere) raises concerns about atmospheric contamination.
Dr. Henke states:
The context of my statement is a catastrophic situation with many volcano eruptions in a short time, and large quantities of argon 40 escaping into the atmosphere. For short periods of time, there could be a significantly larger argon 40 concentration near the ground than elsewhere, making it harder for argon 40 to leave lava as it cools, and possibly increasing the argon 40 concentrations in other rocks.Dr. Plaisted (p. 14) even suggests that argon concentrations in the atmosphere would concentrate near the ground. Of course, such a layer would not form because of atmospheric mixing by winds and because the atomic weight of argon is not that much greater than N2 or O2, the dominant gases in air.
Dr. Henke states:
It wasn't clear to me whether this was based on evidence or conjecture. Argon can travel through fractures, and also be transported by water. It's also not clear to me how well we understand the structure of fractures and cracks in the crust, to understand how argon will circulate. Furthermore, argon may take the path of least resistance, but in some cases it could be trapped underground and spread out into a wide area. Also, if the sediments were hot enough in places, argon could diffuse through them fairly easily, as mentioned above.Young (1982, p. 100-101) refutes Slusher's claims of crustal contamination by argon gas. Because argon is inert, the gas would tend to diffuse through large fractures in the crust rather than into minerals or small fractures within the minerals. That is, the argon would take the path of least resistance.
Dr. Henke states:
However, an article by Snelling at The Institute for Creation Research site lists quite a few historic volcanoes with excessively large K-Ar dates. This is proof that the magma contains large amounts of excess argon, which in these cases was not able to escape.Dalrymple (1984, p. 81-82; 1991, p.91-92), on the other hand, notes that historically erupted volcanics rarely contain excess argon.
Dr. Henke states:
It is difficult for me to find the statements to which Dr. Henke is referring, since my copy of the article does not have his pagination. I don't recall the statements to which he refers.Dr. Plaisted (p. 2, 13) might claim that the excess 40Ar could have formed from increases in the 40K decay rate in the dense and hot interiors of stars. This suggestion is not supported by isochrons (as Dr. Plaisted suggests on p. 13), meteorite chemistry, or the chemistry of the stellar atmosphere.
Dr. Henke states:
Dalrymple (1991) says nothing about this, but in any event, my main concern is not the age of the earth. Dalrymple (1984, p. 83) says that the amount of argon 40 in the atmosphere could have been produced in 4.5 billion years if there are 170 ppm of potassium in the earth, a reasonable amount. Dalrymple assumes that half of the radiogenic argon 40 would be released into the atmosphere, which seems to be an arbitrary assumption. Perhaps there is much more argon 40 in the mantle than in the atmosphere. Slusher (p. 39) cites (Cook, Melvin A., Prehistory and Earth Models, London: Max Parrish and Co. Ltd., 1966, p. 67) in support of his statement. It is also possible that if the speed of light were faster in the past, the decay rate might have been faster as well, producing much argon 40.Dr. Plaisted (p. 10-11) cites an anonymous friend (whose statement turns out to come word for word from Slusher, 1981, p. 39) who claims that there is too much 40Ar in the atmosphere for the Earth to be 4.5 billion years old.
The whole question of how much argon 40 and argon 36 there is in the atmosphere and in the mantle, is discussed in detail in Dickin (1995, pp. 287-293). Faure (1986, pp. 79-80) also discusses this issue, and mentions some work supporting the idea that the mantle lost 95 percent of its argon in a catastrophic event about 4.45 billion years ago. This raises the question as to what kind of an event this could be, especially since gravity would tend to attract argon to the earth, and argon is highly soluble in molten rock. If one assumes that most of the argon 36 present when the earth formed was in the mantle, and that this argon 36 either left the mantle at a constant rate or that most of it is still there, then it must be the case that the amount of argon 40 in the mantle is many times larger than in the atmosphere. These natural assumptions would justify Slusher's statement that there is far too much argon 40 to have formed by radioactive decay from potassium in 4.5 billion years. This could be evidence that the rate of decay was much larger in the past.
Dr. Henke states:
The quote from Evernden et al. (1964) stated:In another example from a web site, Dr. Plaisted (p. 20-21) quotes Evernden et al. (1964), which states that some devitrified (altered) glasses of known ages gave results that were too young. Both chlorites and devitrified glasses are alteration products. Therefore, unreasonable K/Ar dates are expected with these materials because the alteration events would have caused the argon to move in or out of the materials. Obviously, geochronologists would avoid these materials if they wanted quantitative dates.
Thus the evidence is that the devitrification took place at about the same time as the deposit, which contradicts Henke's scenario.Some gave virtually zero ages, although the geologic evidence suggested that devitrification took place shortly after the formation of a deposit.
Dr. Henke further states:
However, Dr. Henke is not giving the whole story here. It follows from the mathematical properties of isochrons that if an isochron, even with carefully selected samples, gives a Rb/Sr age of 34 billion years, then at least one of the samples must have a Rb/Sr age of 34 billion years or larger, and probably at least one more has a Rb/Sr age nearly this large or larger. In fact, from Dalrymple (1984, p. 79), three of the 8 samples have Rb/Sr ages of nearly 34 billion years or larger. Thus we still have a serious anomaly to contend with. Whether this large age is due to leakage of Rb, excess Sr, a different decay mechanism, or a faster decay rate, the same factor could be making other Rb/Sr dates too large, as well. Of course, the first two of these processes probably would not produce false isochrons.Among the citations of Woodmorappe (1979), Dr. Plaisted (p. 17) refers to a ridiculously old Rb/Sr "date of 34 billion years." Dalrymple (1984, p. 77f) discusses the origin of this "date" and denounces Woodmorappe (1979) for creating this fictitious date by misreading an isochron plot for the Pahrump Group Diabase in California (see the original source: Faure and Powell, 1972). The diabase shows a terrible scatter on the 87Sr/86Sr versus 87Rb/86Sr plot and does not provide a Rb/Sr date. Radiometric dating on related rocks, however, indicates that the diabase is about 1.2 billion years old. On the badly scattered diagram, two age-meaningless "reference isochrons" are drawn to bracket the badly scattered data. One reference isochron is "1.09 billion years." The other is a "34 billion years." These reference isochrons have no time meaning. They are simply drawn in as a guide for the reader in much the same way that a flying pilot may refer to the position of another plane as being between "9 and 11 O'Clock" from her position. In this case, the "9 to 11 O'Clock" has no time meaning, but simply indicates that the other plane is from the left to the front left of the pilot.
Dr. Henke states:
What I am looking for is a study in which samples are chosen in a statistically unbiased manner and sent to laboratories without information about where they are found, to determine whether the dates are in acceptable ranges and whether different dating methods agree. Just to give one example does not settle the issue.In Lanphere and Dalrymple (1965), 55 laboratories were sent a muscovite standard for dating. The average K/Ar date for the muscovite was 83.0 million years and the average Rb/Sr date was reasonably close at 85.7 million years. Interlaboratory standard deviations were only 1.2% for the K/Ar dates and 2.8% for the Rb/Sr dates. These excellent results refute creationist claims that K/Ar and Rb/Sr methods are inconsistent or imprecise.
Concerning the dating of meteorites, Henke states:
This does not answer my question, which referred at least in part to dates obtained by a simple daughter-to-parent ratio. Dalrymple (1991) does say that many ages for meteorites are "model ages," which are computed by making assumptions about initial amounts of daughter product. Such assumptions are also necessary for the Pb-Pb method of dating. For meteorites, there is a good basis for making such assumptions. Also, Dalrymple (1991) gives impressive agreements between different isochron methods on meteorites, which support a roughly 4.5 billion year age without any assumptions on initial amounts of daughter product. Henke refers to this in his second reply:Dalrymple (1984, p. 82) states that knowledge of the initial amount of the daughter product is not even needed to obtain reliable ages with isochron methods.
Other than a true age for the meteorites, this could indicate a mixing process, which does not seem likely, or an increase in decay rates, for which a mechanism would need to be found. However, in order to date the earth, one needs an isochron which includes a point from the earth. This is more difficult, and this is the isochron that I saw on the talk.origins FAQ, which involved a much smaller number of meteorites. Also, many of the meteorite dates I saw in the FAQ were apparently simple daughter-to-parent ratio ages.Dalrymple (1991) lists more than just a "small number of meteorites." For example, tables in Dalrymple (1991, p. 287-289, 291) list numerous results. Specifically, Figure 6.9 on p. 286 contains the results of 94 radiometric ages of 69 meteorites. Table 6.3 on p. 287-289 contains the results of 240 radiometric dates on 42 meteorites. None of these results are only a few thousand years old. For further details, see Dalrymple (1991, chapter 6).
It is also remarkable that so many different isochron-based dating schemes, even on the same meteorite, often yield roughly the same 4.5 billion year age. This is true for Rb-Sr, Pb-Pb, Sm-Nd, Lu-Hf, and Re-Os dating methods, as well as the Ar-Ar spectrum method. This is a case where different methods agree without making assumptions about initial amount of daughter product. This either indicates a true age, or a change in the decay constants. I would like to know how often this is true on the phanerozoic. How often does one have two or three different isochrons on the same system yielding very similar dates? It is also important that the concentrations of parent substances are linearly independent, to preclude mixings. Such a multiple-isochron agreement is fairly convincing, but the failure to find such isochrons likewise casts doubt on the ages obtained. If radiometric dating is accurate on fossil-bearing rocks, there should be an abundance of such agreements between different isochrons on the same systems, and they should yield the conventionally accepted ages. A change in the decay constants on the phanerozoic seems less likely, since it could radically affect the properties of matter, and be harmful to life.
Another reliable technique mentioned by Dalrymple (1984) is the U-Pb concordia-discordia method on zircons, which is valid even for many open systems. This technique requires assumptions about lead and uranium loss, and seems to give good evidence of a reliable date (relative to decay constants), especially when there is agreement with other methods (such as isochrons) on the same system. However, Dickin (1995, pp. 105-118) mentions that both lead and uranium are mobile, and that "rocks with complex geological histories ... can yield discordia of high statistical quality which nevertheless yield erroneous ages" (p. 118). Also, zircons can survive transport through magma (p. 116-7) and become damaged by decay of uranium, permitting the loss and gain of lead from the environment (p. 108-9). Woodmorappe (1999, p. 33) gives a number of references showing that inherited zircons, and inherited lead in zircons, are common, and give ages that are older than the accepted ages. This leaves open the possibility that all of the dates obtained from zircons on the phanerozoic are too old.
I mentioned the presence of excess argon 40 in a sample as a problem
leading to artificially old K-Ar dates. Henke states in a reply to
me, concerning the problem of detecting excess argon,
It is possible that such isochrons are not often done. One cannot always use an isochron, since many minerals may have about the same K and Ar40 concentrations, and there may be some fractionation of argon among the minerals. It's not clear to me if this three dimensional plot always works, and how often it is used. I was not able to find any mention of it in Faure (1986) or Dickin (1995). Dickin (1995, p. 116) did mention such a technique for U-Pb dating.Why not determine the initial Ar isotopes with K/Ar isochron methods? Also see Mussett and McCormack (1978) on using a three dimensional plot to distinguish initial and excess argon in K/Ar dating.
Henke further states in his reply to me,
It is true that by using additional isotopes (if they are sufficiently abundant and do not fractionate), one can often detect mixings of multiple sources. My point was that the usual mixing test can only detect two sources. But since these multiple mixing tests are more difficult and expensive, they may not be done very often. One also has to know which isotopes to examine. I was suprised that Dalrymple (1991) said nothing about mixings invalidating isochrons. Dalrymple (1984, p. 85) cites a source (Kramer, Arndts, and Overn, Bible-Science Newsletter, 1981) that applied the mixing test to 18 Rb-Sr isochrons from the literature and found that nearly all of them had correlations suggesting a mixing. Dalrymple goes to great lengths to explain this away, but I think this figure is very telling, and find his explanations unconvincing. It is also remarkable that we have a test for mixing, which is commonly cited in support of the accuracy of radiometric dating, but when it gives contrary results, it is simply ignored. Dickin (1995, p. 44) states,Dickin (1995, p. 180-199) mentions a number of other isotopic techniques that are used to detect mixing between multiple sources.
Dickin suggests that mixings may contribute to such isochrons. It seems reasonable, then, that mixings may be affecting all Rb-Sr isochrons in igneous rock.It is a fundamental assumption of the mantle isochron model that neither isotope nor elemental ratios are perturbed during magma ascent through the crust. However, it is now generally accepted that this assumption is not upheld with sufficient reliability to attribute age significance to erupted isochrons.
Henke states:
This example is given to show that a mixing of three sources cannot be detected by the usual two sources test. It is not intended to be natural, but to demonstrate a mathematical fact. There is a lot of flexibility in the design of such examples, as I indicate, and it is reasonable to assume that some of these examples would be natural. It's the responsibility of the geologist to show that such mixings have not occurred.Your hypothetical example in "More Bad News for Radiometric Dating" is often hard to follow, but it is clearly invalid.
In response to the disagreements between different dating methods, Henke states:
The problem with this approach is that it leaves ample room for the exercise of subjective judgment and evolutionary assumptions. Also, Dalrymple (1991) says essentially nothing about the phanerozoic, and thus gives little evidence of the accuracy of the conventional dating scheme on fossil-bearing rocks. Dalrymple (1984, p. 125) also says little about the phanerozoic, except to note that over 100,000 dates have been calculated, and only a few percent (p. 76) are anomalous. I treated this issue of percentage of anomalies in considerable detail in my original "Radiometric Dating Game" article.To really understand what's going on you have to sample the recent works of many different authors. You have to follow arguments between experts on different issues and see where they go. Overall, the geologic time scale is in great shape. Yes, scientists are still making minor adjustments. However, it's clear from Strahler (1987), Dalrymple (1991), etc. that the creationists have lost.
It is interesting that Woodmorappe (1979) gives a number of cases in which standard geological tests are ignored. For example, dates may be accepted even when there is evidence of weathering, and rejected when there is not. There may be evidence of heating, but the date may be accepted, and there may be no such evidence, but a hypothetical heating event is assumed anyway. The same applies to the Ar/Ar spectrum test for initial or adsorbed argon 40. If geological tests are not being applied consistently, one wonders what value they have.
Let me clarify the problem with excess argon. There is an excellent discussion of radiometric dating from Cornell at Tim Thompson's radiometric dating page. It gives the diffusion equation for argon escaping from a rock as it cools. The rate of diffusion is proportional to the gradient of argon concentration, and increases rapidly with temperature. Suppose the partial pressure of argon 40 in the environment is p. Suppose the partial pressure of argon 40 in lava or magma is initially at least p, as it cools. Then the partial pressure of argon 40 in the magma will never decrease below p; excess argon 40 will remain dissolved in the lava or magma as it cools. This argon 40 will then be trapped within the resulting rocks and lead to artificially old K-Ar dates. Now, the problem with this is that this excess argon 40 will probably be deposited as single atoms of argon distributed evenly within the sample. This makes it very difficult or even theoretically impossible to distinguish this excess argon 40 from argon generated by radioactive decay. This will make the sample appear artificially old right away. Even if crystals exclude argon as they form, argon will rapidly diffuse into them as the lava cools, by the diffusion equation mentioned above.
A similar problem can occur if the excess argon 40 dissolved within lava or magma is not able to escape, due to rapid cooling or subsequent deposits of sediment or other lava on top.
It is possible that in some cases an isochron might be able to detect such initial argon 40, but this can only happen if the potassium concentration varies significantly within the sample. It is not clear to me, also, how often such a test for initial argon 40 is performed. And of course, such isochrons can be falsified by mixings or other problems.
There are spectrum tests for adsorbed argon involving Ar-Ar dating; basically, one can see whether the argon 40 is concentrated near the surface of the sample or near the interior. The former would indicated adsorbed argon 40, which would not give a true age. However, this test would not indicate excess argon 40 present during cooling. Faure (1986, p. 104) verifies that biotites with excess argon can have a flat spectrum, which makes this excess argon impossible to detect. Faure (p. 101) even implies that the reliability of the Ar-Ar spectrum test is now being questioned.
It seems reasonable to me that this is a uniform problem with K-Ar dating. To me the geological evidence suggests catastrophic conditions and rapid formation of the sedimentary layers in the past. Thus the lava might have been covered before the excess argon was able to escape. Or the lava might have cooled quickly, due to rainfall. It only needs to cool to about 500 degrees centigrade or less to trap most of the argon, at least for biotite.
As I mentioned before, one sometimes finds significant argon 40 in a rock and no potassium at all, as mentioned in Snelling's article. This shows that excess argon is entering these rocks by some means, and calls K-Ar dating into question.
Excess argon could even cause different minerals in a given formation to yield similar K-Ar ages, since they all might have similar concentrations of K, approximately equal to its abundance in the earth's crust, and similar concentrations of argon 40, due to the partial pressure of argon 40 being similar during cooling. Even sedimentary minerals might have a similar K-Ar age for the same reason. Also, lava (magma) that cooled within the earth is likely to have artificially old K-Ar ages, since the enclosed excess argon 40 might have a more difficult time escaping.
One sedimentary mineral of particular importance for K-Ar dating is glaucony. The following message from a talk.origins participant illustrates its importance:
Concerning glauconite, Dr. Henke saysFor example, Plaisted's "explanation" for the correlation of isotopic age with vertical position in the geologic column is essentially that excess argon would have existed in lavas in greater quantity early in the Flood, and decreased as it was outgassed over time. Had Plaisted actually bothered to look at the data (e.g., Harland et al. "A Geologic Time Scale 1989") he would have found that more than half of the dates, especially for mesozoic-cenozoic parts of the column, are glaucony (a sediment composed largely of animal fecal matter). Glaucony did not come from a "magma chamber," so Plaisted's explanation cannot possibly cover the majority of ages on the younger parts of the column.
The fact that glauconies are unreliable is significant, since they provide such a large part of the dates for the mesozoic-cenozoic parts of the geological column. Faure (1996, p. 71) also lists glauconies as being only "sometimes useful" for K-Ar dating.Of the 429 or so "anomalous" dates in Woodmorappe (1979), 94 (21.9%) involve illite and glauconite, which (because of their crystalline structures) are notoriously unreliable in giving K-Ar dates. Geochronologists know that illite and glauconite K/Ar dates are often untrustworthy. K/Ar ages from these minerals are often published to better understand the types of conditions that cause them to produce unreliable dates rather than to assign actual ages. Woodmorappe is clearly misusing illite and glauconite dates to simply pad his list.
Glauconies are formed in seawater from a variety of materials, and incorporate potassium from the seawater (Faure, 1986, p. 78). Faure states, "The uncertainty about the interpretation of K-Ar dates of `glauconite' has reduced their usefulness in dating sedimentary rocks" (page 78). The process of their formation gives a ready mechanism for their K-Ar ages, namely, the incorporation of argon 40 as well as potassium from the seawater. We can assume that as a result of a global catastrophe, the oceans were highly enriched in argon 40 in the past, and that the concentration of argon 40 gradually decreased over time, due to its diffusion into the atmosphere and due to a smaller amount being released into the seawater. Therefore older glauconies would absorb more argon 40 from the seawater, resulting in old K-Ar dates for lower strata which become progressively younger for higher strata. Another factor in this direction is that older glauconies have more time to absorb argon 40.
Let me summarize some of the reasons for questioning the reliability of K-Ar dating:
Finally, I want to comment on the circumstances of the interchange with Dr. Henke. During most of our interchange, I was not aware that it would be published on talk.origins. Now it has been web-immortalized on a radiometric dating web page. I was not informed that this exchange had been posted there. In addition, the complete exchange was not posted, but only a portion of it. I do thank Tim Thomson for the courteous and professional manner in which he has interacted with me, and that he has included the rest of my exchange with Dr. Henke.It is difficult to detect argon 40 present during cooling.
Some minerals contain argon 40 but no potassium, so this indicates excess argon 40, which in the presence of potassium leads to artificially old dates.
Many historical volcanoes give K-Ar dates that are much too old, even if the reasons for this are understood.
At least one authority sent me email that K-Ar dating is inaccurate.
Fission track dating is an independent dating technique that needs to
be evaluated by creationists. This method is based on observing the
fission tracks produced by fission of U 238 in a sample. These tracks
are observed by immersing the mineral in an acid solution so that
the tracks are etched away and become more visible. By knowing how
much U 238 is present and how many tracks are produced, one can
compute the age.
It is interesting that Faure (1986, pp. 345-6) mentions that fission track dating is calibrated (the "zeta calibration") using rocks of "known" ages. If these "known" ages are incorrect, then fission track dating is also incorrect. It appears that this "zeta calibration" could be introducing errors into fission track dating, because it essentially uses the "known" date of a standard material to estimate how much U 238 this sample has, instead of measuring the U 238 concentration directly. For example, an error in estimating the U 238 concentration would imply a similar error in the measured age of the sample. The standard sample is typically a zircon, which has its own problems of dating. This calibration also assumes a constant ratio of U 235 to U 238.
The principal method of decay of U 238 is by the emission of an alpha particle (a helium nucleus), and this occurs more than a million times as often as fission. This causes the U 238 nucleus to recoil, which generates a small track. Not only this, but the resulting nucleus decays again, and all in all there is a chain of 14 decays involving ejection of an alpha particle or beta particle (an energetic electron) before lead is reached at the end. Thus we have over 14 million other particles being emitted and potentially damaging the sample for every fission track, not to mention the gamma radiation emitted (high energy electromagnetic radiation). Over long periods of time, these other particles and radiation might cause significant damage, and perhaps create structures approximating fission tracks.
The way that alpha particles damage crystal structure is that their positive charge attracts electrons, so as an alpha particle passes by an atom, some of the electrons may be attracted to the alpha particle and flung out of the atom. This breaks chemical bonds and damages the crystal structure. It also creates positive ions which repel each other, and this repulsive force causes stress on the crystal structure. Alpha particles also create pleochroic haloes around radioactive inclusions in rocks by this process. In addition, they cause the crystal structure of zircon to become damaged, or "metamict," over time. Although alpha particles are too small to create fission tracks, they do create trails of damage in the crystal. Perhaps if two alpha particles create trails that are in nearly the same location, especially when the substance is already damaged by radiation, these trails can contain enough damage to the crystal so that when it is annealed with acid, a track-like structure results.
Another problem with fission track dating is that the tracks "anneal," or gradually disappear, with time. This implies that old rocks will have many fewer tracks, and shorter tracks, than one would expect if all tracks were preserved. However, the small number of tracks could also be an evidence that these "old" rocks are younger than geologists assume. This process of annealing occurs faster at higher temperatures. It is necessary to extrapolate this process of annealing to large time intervals in order to use fission track dating, and this extrapolation has many uncertainties. Dickin (1995, p. 425) says, "In the above discussion, geologically well-known thermal basin histories were used to calibrate the annealing behavior of apatite tracks." Of course, if these histories are in error, this calibration introduces another significant source of error in fission track dating. The lack of knowledge of the temperature history of rocks is another source of possible error.
All of the fission track dates mentioned in Harland et al (1990) for calibrating the geological column are on zircons. Zircons are known to survive transport through lava, so these fission track dates could be in error, and older than the dates of the layer they are found in. According to Dickin (1995, p. 422), about half of the fisson tracks in sphene disappear in 100 minutes at about 600 degrees centigrade. Zircons have a slightly lower closure temperature than sphene for fission tracks, so it is conceivable that many fission tracks could survive transport through lava or magma and lead to anomalously old ages.
Finally, it is possible for uranium to be leached from a sample. If 9/10 of the uranium has been removed, then the fission track date will be ten times too large. All of these factors put together cast considerable doubt on the accuracy of fission track dating.
Let me briefly comment on a couple of other articles at Tim Thompson's
page. One by Don Lindsay,
Are Radioactive Dating Methods Consistent With Each Other? shows
that 5 craters give similar dates, three of them dated by three
different radiometric methods, and two by stratigraphic ones. This is
at least close to what I am looking for. However, it would be better
to date all five craters by all four different methods, and see what
the agreement is. It is also possible that each crater gives a
scatter of dates, and the best ones were selected. Furthermore, it is
possible that the craters were chosen as those for which the dating
methods agreed.
I also comment on the article Breakthrough Made in Dating of the Geological Record at the same site. This article shows an agreement between argon-argon dating and astronomical time scales which is used to calibrate sedimentary deposits. This is very interesting, but the assumption is that the precession of the earth's axis has been constant for many thousands of years. If there were a recent global catastrophe, this precession might have been severely altered, invalidating this calibration. In general, uniformitarian assumptions such as this were foreseen in the Bible:
And Dalrymple (1984, p. 104) even defines uniformity, following Hubbert (1967, p. 31), byKnowing this first, that there shall come in the last days scoffers, walking after their own lusts, and saying, Where is the promise of his coming? for since the fathers fell asleep, all things continue as they were from the beginning of the creation. For this they are willingly ignorant of, that by the word of God the heavens were of old, and the earth standing out of the water and in the water; whereby the world that then was, being overflowed with water, perished.
2 Peter 3: 3-6.
However, the assumption of global catastrophes in the past calls such calibrations of dating methods into question.(1) We assume that natural laws are invariant with time
(2) We exclude hypotheses of the violation of natural laws by Divine Providence, or other forms of supernaturalism.