Tektites and Their Origin - Chapter Eight

 where ρ_c  is the density of the cloud, and ρ₀ is the density of a mass equal to the mass of the sun, distributed over a sphere whose radius is the distance from the sun to the cloud.  Numerically ρ_c = 2.8 x 10^-4 kg/m³.  If the tektites of the Australasian strewn field had ever moved about in space they would presumably have formed a cloud whose diameter would be of the same order as the Australasian strewn field, namely 8800 km.  With the above density, the mass in a column going through the cloud would be   2.4 x 10⁵ kg/m³.  When the cloud struck the earth, it could be expected to produce a surface deposit of this density.  The actual surface density in the Australasian strewn field is estimated by Glass (1972a) as 10^-3 kg/m³; and this estimate is free from the objections about losses by aboriginal man, stream action, birds, mechanical erosion, fusion stripping, sand-blasting, rolling, and probably even solution etching, raised by Baker (1960a), since Glass's estimate depends almost wholly on the density of microtektites at the bottom of the sea.

Hence a cloud of tektites arriving from space would necessarily have had too low a density for stability.  Then, unless the space trajectory was very short, the tektites would have formed a cloud with dimensions greater than the earth.  It follows that, instead of the observed strewn fields, there would have been a distribution over a whole hemisphere of the earth, as is observed in meteor showers.  Since this is not what happens, it is generally agreed that origins more remote than the moon can be excluded.

Chapman (1964, 1971) showed that the existing strewn fields are reconcilable with origin from the moon, provided that the tektites come directly from the moon.  Urey (1963) referred to earlier versions of this idea, and made the point that if tektites are from the moon, then in any burst of tektites, the great majority ought to miss the earth and go into heliocentric orbit.  They would then eventually strike the earth, and would form a conspicuous population of tektites distributed at random over the earth, which is not observed; and he concluded that they must be terrestrial.

O'Keefe (1963) and O'Keefe and Shute (1963) made an abortive attempt to meet Urey's objection by supposing that tektites enter the atmosphere from decaying, nearly circular orbits, derived from a single, large parent body for each strewn field.  The large parent body was an essential party of the theory; with more than one parent body, there would be a range of decaying orbits, and hence a distribution completely around the earth.  The parent-body idea had to be given up (O'Keefe, 1969a) when it became clear that the microtektites could not be explained as ablation drops from the same parent body as the large tektites; and hence that they were apparently independently launched from the same source as the macrotektites and at the same time.

It thus follows that the terrestrial origin of tektites must be the true one unless there is some mechanism in space which prevents tektites, once launched into heliocentric space, from being recovered by the earth.  Urey (1963) pointed out that the Poynting-Robertson effect will require about 10 million years to destroy a particle of 1 cm radius, while capture by the earth will, he calculated, require only a few times  10⁵ years; similarly space erosion at the expected rate of about 1 nm per year is too slow.

Radzievskiy (1954) showed (in a different connection) that such a mechanism does exist.  He found that radiation pressure will exert a torque on a body if the albedo on one side differs systematically from that on the other (as in a Crookes radiometer).  The two sides involved here are not the front and back, for instance, but the left and right sides as viewed from the sun (Fig. 41).  He took as an example a 1-cm cube, rotating around a vertical axis perpendicular to the direction of radiation.  If, on each of the vertical faces, the left side is black, and the right is white, then the cube will rotate in the counter-clockwise direction.  In sunlight, the cube will spin up to bursting speed in a few times  10⁶ years if the albedo difference is systematically as much as about 1% and if as much as 1% of the light on both sides acts like a specular reflection.

Paddack (1969, 1973) further showed that an object of helical shape will be set into rotation by light directed along its axis (Fig. 41).  He found that there is a significant helical component in the external shapes of ordinary pebbles.  Like the Radzievskiy effect, the Paddack effect leads to ages which are orders of magnitude less than the Poynting-Robertson effect; hence Urey's argument against a cosmic origin for tektites cannot be extended to exclude a lunar origin.

 The isotopic studies on  ²⁶Al (Viste and Anders, 1962) and ¹⁰Be (Rayudu, 1963) (Chapter 7) indicated an upper limit of about 10⁵ years for the residence-time of the Australasian tektites in space; and a narrower limit of less than  10³ years was set by fission-track studies (Fleischer et al., 1965a).

An extension of Urey's argument was made by Barnes (1958a) and Kopal (1958) who argued that the structure within a given strewn field is so fine that it is difficult to believe that tektites originate at the moon.  For example, the Bohemian and Moravian portions of the moldavite strewn fields are each only about 50 km in width.  Similarly McColl and Williams (1970) find that there is a high concentration of australites in long narrow bands less than 100 km in width.

This fine structure can perhaps be understood, even in terms of a lunar origin, if there is a gas accompanying the tektites at the source region. The gas will entrain the tektites, so that all tektites in a given region will move with the same velocity, namely the gas velocity.  Mixtures of this kind are well known in studies of atomic explosions; they are called ensembles.  (Note that if the gas is at rest in a gravitational field, the velocities of the solid particles contained in it will tend toward a terminal velocity, which, for the case of the earth, is a few tens to hundreds of meters per second; but if the gas is moving on a ballistic trajectory, then the velocities of the solid particles will tend toward zero with respect to the gas, and therefore with respect to each other.)  If the relative velocities are reduced to a few tens of centimeters per second, then, even in the passage from the moon to the earth, the fine structure will not be smeared out at the level of tens of kilometers.  The gas will have cooled as a result of expansion; this will reduce its tendency to expand further  A quantitative treatment would be valuable, taking into account the focussing effect of the earth's field.

CHEMICAL AND ISOTOPIC ARGUMENTS:  S. R. TAYLOR'S REVIEW

The strongest arguments for the terrestrial origin of tektites, and against the lunar origin, are in the field of chemical and isotopic studies.  These arguments have been ably reviewed by S. R. Taylor (1973), using Apollo results.  In this section, Taylor's arguments for his theories will be stated, point by point, omitting those points which do not bear on the question of terrestrial versus lunar origin.  Each point will be followed by a comment (of mine).  The headings are from his paper.

Chemistry

Point:  Tektites and impactites seem to favor the high-silica end of the petrological sequence; SiO₂ less than 65% is rare.
- Comment:  A large fraction of the microtektites, though not the majority, have SiO₂ less than 65% (Cassidy et al., 1969); the Lonar Lake impact glasses are around 50% SiO₂ (Fredriksson et al., 1973).  Hence there are no physical barriers to low-silica tektites or impactites.

Point:  There is a narrow range in tektite composition; it suggests that tektites are the result of a process of natural selection; somehow rocks outside this range fail to form tektite glass.

- Comment:  The most important parameter governing the physical performance of tektite glass is undoubtedly silica content, especially through its influence on viscosity.  But tektites cover a wider range of silica content than most types of terrestrial igenous rocks, namely from 49 to 98%; this results in a range in viscosity which has never been measured, but is certainly more than six orders of magnitude.

If Taylor's hypothesis were right, we would expect tektites to show a relatively small range of silica content, and a wide range in other oxides.  But in fact the range in silica is comparatively wide, and that in other oxides, at a given silica content, remarkably narrow, especially for a given tektite clan.

Point:  Aouelloul glass is identical with Zli sandstone, according to Cressy et al. (1972).
- Comment:  Cressy et al. measured five elements and the strontium isotope ratios; the agreement was good, but there are other elements for which agreement is not good, including calcium, iron (O'Keefe, 1971), and a series of trace elements (C. S. Annell, unpublished).  The glass has much lower water content and ferric/ferrous ratio (see pp. 137-139 [L.Hancock note:  for pp. 137-139 in 1976 edition, see in this version Chapter Five, section "Aouelloul crater glass versus local {Zli} sandstone"]).

Point:  Australites are chemically like Henbury impact glass (i.e. well-studied tektites often resemble well-studied sandstones).
- Comment:  The resemblance does not include water, ferric/ferrous ratio, CO₂, or the relatively volatile elements on the right side of the periodic table (see Fig. 31, and p. 120 [L.Hancock note:  for p. 120 in 1976 edition, see herein Chapter 6, the section "Central Composition"]).  

Point:  The zircon-baddleyite reaction fixes the temperature of formation of tektite glass at around  1900° C .
- Comment:  In hot glass, this reaction takes place at temperatures which need not be higher than 1500° C (see p. 94 [L.Hancock note:  for p. 94 in 1976 edition, see herein Chapter 4, the section "Zircon and baddleyite"]).

Significance of lunar data

Point:  Lunar glasses do not resemble tektites in chemical composition.
- Comment:  Most lunar glasses are basaltic.  Basaltic volcanism is typically of the Hawaiian type, characterized by effusive flows, rather than the explosive type needed to propel tektites to the earth.  Some sialic glasses from the lunar soil are not unlike tektites (Table XI).

Lunar sample 14425 is an 8-mm glassy sphere, not yet analyzed, with a bulk specific gravity of 2.6, like certain javanites.

Point:  Lunar rocks have higher chromium values (up to 0.25%).
- Comment:  Similar chromium values are seen in bottle-green microtektites (Glass 1972b).

Point:  Lunar rocks are generally richer in the refractory elements (Ti, Zr, Hf, and the rare earths).
- Comment:  Both lunar samples, as returned by Apollo and Luna projects, and tektites have been through a baking-out process, but it was more thorough for the material of the returned lunar samples than for the tektites.  The difference may be a question of depth in the moon; the outermost parts may be poorer in volatiles.  

Point:  Lunar basalts have a peculiar pattern of rare earth elements, not seen in tektites.
- Comment:  Apollo 16 anorthositic gabbros have a rare earth pattern which is very different from that of a mare basalt, and, in some cases, much like tektites (Taylor et al., 1972).  

Point:  Lunar rocks have K/U ratios near 1000-2000; terrestrial rocks and tektites have K/U ratios about 10,000.
- Comment:  Lovering and Wark (1975) find K/U ratios averaging 6000 and going up to 10,000 in lunar granites and monzonites.  Some of these resemble tektites chemically.  Metzger et al. (1973) find a variable ratio over the moon (by remote-sensing techniques); the ratio is particularly high on the far side, though it does not reach 10,000.  

Point:  Tektites could not be produced from lunar mare rocks by selective distillation.
- Comment:  This rules out origin by impact on mare rocks, as earlier suggested by me (O'Keefe, 1963).  

Point:  Lunar basalts of the type called KREEP (rich in K, the rare earth elements, and P) are not like australites in chemical composition.
- Comment:  They are much more like the bottle-green microtektites (see Chapter 6, pp. 139 and 140 [typist's note:  See Chapter 6, the section "12013 compared with the high-magnesium tektites"]).  

Point:  Rock 12013, whose major element chemistry is like certain javanites, is not glassy as stated by O'Keefe (1970).
- Comment:  True.  

Point:  Discrepancies in the trace element composition of 12013 (as compared with tektites) are so serious that 12013 is not related to tektites.
- Comment:  Each new group of tektites to date has differed from the previously accepted definition of tektites in some minor way.  Chapman (1971) thinks that 12013 belongs among the tektites, and particularly with the high-magnesium clan.

Point:  Lunar sialic (granitic) rocks are generally higher by factors of 10-30 in K/Mg, and K/Na, compared to australites.
- Comment:  Australites are not the best comparison: moldavites are richer in potassium, and bediasites are much poorer in magnesium.  On the lunar side, Taylor took for his lunar sialic rocks the mesostasis (late-stage residual glass found between the crystals in basalts) or some immiscible globules found by Roedder and Weiblen (1970).  These are indeed richer in K₂O and poorer in MgO than tektites.  But in the lunar soils, there is a glassy component (see pp. 140, 141 and Table XI [typist's note:  for pp. 140-141, go to Chapter 6, the section "Lunar soil particles"]) which does not show the peculiarities of the mesostasis, and is much like tektite glass.  The abundance, in the fines smaller than 1 mm, is often between 1 and 0.1%.

Point:  Ages, especially ages of differentiation, so far found in lunar samples, are not below 3.2 billion years; this suggests that the moon ceased to be volcanic some billions of years ago. Yet tektites have ages of differentiation which do not exceed 2 billion years, and at least some have ages under 50 million years.

- Comment:  Lunar sample 77017.32B (Kirsten and Horn, 1974) yields a good K-Ar age of -1.5 billion years.  Moreover, there is evidence, too complex to be given here, for rare contemporary paroxysms of lunar volcanism (see Chapter 10).  If tektites are propelled by volcanism, their ages of differentiation would be expected to be near the ages of arrival at the earth's surface.  The evidence seems to be reconcilable with this view (Chapter 7).  The ages of arrival at the earth's surface are all relatively recent, because tektites have been found only in unconsolidated surface sediments.

Point:  Lunar leads are radiogenic (high ratios of ²⁰⁶Pb, ²⁰⁷Pb, ²⁰⁸Pb to ²⁰⁴Pb).  These indicate that the moon lost its lead at an early stage, and most of the lead now found there is due to the decay of uranium and thorium.  Tektite leads, by contrast, have ratios about a factor of 10 smaller; and terrestrial leads are like tektite leads.

- Comment:  After the publication of Taylor's paper, it was found that some Apollo 17 leads are non-radiogenic; they are just about what would be expected for terrestrial or tektite leads at the date of the Apollo rocks, about -3.15 billion years.  From this finding, both Tatsumoto et al. (1973) and Silver (1974) concluded that the moon must have somewhere within it regions which contain more volatile elements than the material of most of the Apollo samples.  If tektites are not terrestrial, they point in the same direction, as noted above.

Point:  The oxygen isotope data on tektites does not fit the lunar oxygen isotope data, including that for specimen  12013.

- Comment:  As noted in Chapter 7, pp. 163, 164 [typist's note:  Chapter 7, the section "Oxygen isotopes"], the tektite oxygen isotope data is generally like that of terrestrial acid igneous rock (δ¹⁶O +9 to +11.5 ⁰/₀₀ ).  Lunar oxygen is much like that of terrestrial basalts.  Since 12013 contains a mixture of granitic and basaltic rock, in close proximity, there may have been equilibration.

Tektite oxygen isotope data, except that for the Ivory Coast tektites and Darwin and Aouelloul glass, does not fit that of terrestrial sedimentary rocks.

Possible terrestrial parent material

Point:  Tektites are not at all like any kind of basaltic or intermediate terrestrial rock; and detailed comparison with granites excludes them also.

- Comment:  Tektites are most like intermediate rocks (andesites or dacites, for instance) except for a difference in silica content of the order of 10% (see Chapter 6, pp. 119 and 120 [typist's note:  see Chapter 6, section "Central Composition"]).

Point:  Variations in element ratios among the tektite sequence are not like those observed in terrestrial igneous rock suites.

- Comment:  They can plausibly be explained as igneous suites produced from magmas at low pressures (Chapman and Scheiber, 1969).   

Point:  Tektites are more like terrestrial sedimentary rocks than like terrestrial igneous rocks, especially in the rare earth elements.

- Comment:  In the major elements, it is principally the enhanced silica which produces a resemblance between tektites and certain sandstones.  In the sandstones the cause is the high resistance of quartz to weathering.  Some other cause must be responsible for the enhanced silica in specimen 12013, where it is combined, as in tektites, with a high abundance of mafic oxides.

In the rare earths, the Henbury subgraywacke is only marginally closer to tektites than the standard andesite, AGV-1 (see Figs. 31 and 32). 

Point:  Glass (1970b) found chromite and zircon in Muong Nong tektites; these are very resistant minerals, characteristic of terrestrial sediments.

- Comment:  Finkelman (1973) found zircon and chromite crystals in the ultrafine fraction of lunar soil; this soil also contained about 0.2% of glass resembling microtektites.  Glass later (1970c) found rutile, corundum and monazite in Muong Nong tektites; of these, rutile was also found by Finkelman, and corundum by Kleinman and Ramdohr (1971).  

Point:  Bottle-green microtektites might result from the melting of chlorite in a sediment.

- Comment:  Chlorite is rare in rocks which would yield a recent age of differentiation.  It is characteristic of the metamorphic rocks of the basement complex, typically over 600 million years old.  It contains over 10% water, thus making the problem of the elimination of water one order of magnitude more difficult than for most rocks.  Finally, the extreme bottle-green microtektites would have to start from rock that was 2/3 chlorite; such high concentrations are not typical.

Origin of tektites

Point:  If tektites are produced by volcanic eruptions, then particles of tektite composition should turn up in lunar soils; but they do not.

- Comment:  The compositions of lunar granitic soil particles shown in Table XI look like tektites.  Material of this kind seems to form 0.1-0.5% of the glassy fines of the lunar soil.  In a layer 10 m thick, in which half the material is in the fine fraction (200 µm or smaller), the indicated quantity of tektite-like material is then roughly 10 kg/m²; this is about four orders of magnitude higher than the surface abundance of tektite glass in the Australasian strewn field.  Clearly Taylor's statement is not justified in the present state of our knowledge.  A thorough investigation of these granitic glasses is needed.

Point:  The nickel-iron spherules found in tektites indicate meteorite impact.

- Comment:  The nickel-iron spherules have been found only in philippinites (Chao et al., 1962), Dalat tektites (Chao et al., 1964), and Aouelloul glass (Chao et al., 1966a); this contrasts with their great abundance in true impact glasses.

Within the Australasian strewn field, the nickel abundance averaged 9.1% in nickel-iron from Dalat (Viet Nam), but 2.9% in nickel-iron from Isabela in the Philippines.  If there was a single impacting body, this result is unexpected.

The value of 2.9% nickel is exceptionally low for meteoritic iron; but Brett (1967) points out that in terrestrial impact glasses, nickel is almost always enhanced above the level in the impacting meteorite, because iron oxidizes preferentially and then dissolves in the surrounding silicate.

The nickel-iron spherules are associated preferentially with the low-calcium; high-aluminum tektite composition, both in the Philippines and in the spherule-bearing layer at Aouelloul.

The moon does have both endogenous and meteoritic nickel-iron spherules (Goldstein and Yakowitz, 1971) which do show a wide variation in nickel abundance; values below about 4% are classified as endogenous by these authors.

Point:  Coesite (Walter, 1965) indicates meteorite impact.

- Comment:  The finding of coesite was difficult and should be confirmed.  It does not seem to be present in all Muong Nong tektites.  The peak shock pressures at which coesite forms correspond to Stage II and Stage III of Von Engelhardt and Stöffler (1968).  At these stages, unmelted feldspar is encountered, as well as melted grains which have not become mixed.  Neither feldspar nor glass of feldspar composition has been reported from Muong Nong tektites (except in very small quantities (Chantret et al., 1967) in the lechatelierite), although it is a necessary component of the hypothetical parent sandstone.  Hence the coesite, if present, cannot have been formed by the same process as that which (according to the hypothesis of the terrestrial origin of tektites) made the glass.

Impacts of sufficient velocity to produce particles moving at the required velocity should have melted the shocked rock, and very probably vaporized it.

Impact sites

Point:  Ivory Coast tektites are established as coming from Bosumtwi crater.

- Comment:  The evidence is not convincing.  The lead is young; Bosumtwi crater rock is old.  The Rb-Sr age fit is forced.  The K-Ar age fit was first announced as occurring at 1.3 million years (Zähringer and Gentner, 1963); this was clearly a coincidence since the same authors later supported ages of 1.0 million years for the Bosumtwi and the Ivory Coast tektites; the possibility exists that the present agreement is also illusory.  The chemical comparison is not precise; it is much improved if only Ata glass is used; but in this case the oxygen isotopes do not agree.  No compositions corresponding to the bottle-green microtektites have been reported around Bosumtwi.  

Thus it is premature to say that identity has been established between Ivory Coast tektites and Bosumtwi crater rock.

MISCELLANEOUS ARGUMENTS

Gases in tektite vesicles

Müller and Gentner (1968) and Jessberger and Gentner (1972) have studied the composition of gases in tektites of the Muong Nong type. They find that the nitrogen and the rare gases in the vesicles of these tektites are unmistakably atmospheric (i.e. similar to the earth's atmosphere) in their abundance relations to each other and in the relative isotope abundances in each gas.  They further find that as compared with the earth's atmosphere, the CO₂  is greatly enhanced (from the atmospheric value of about 0.03% to around 50% or more); while oxygen is depleted to values near zero.  They conclude that the gas cannot be the result of leakage into the vesicles after the tektite solidified. They consider that the gas must have been trapped during an impact, when the oxygen had been consumed by fires, and the CO₂ had been released by the calcining of limestone.

This explanation neglects the kinetics of the problem.  It is not possible to burn up a substantial log of wood in the second or two of a meteorite impact, as a calculation of the thermal diffusivity will show and as common sense will confirm.  Hence the only way to produce the CO₂ would seem to be to vaporize the vegetation, along with the rock, by shock.  In the recombination process, the CO₂ might be produced from its elements.  It is suspicious, however, that neither CO nor O₂ was left over; the constituents must have been present in very nearly their stoichiometric concentrations.

It is also curious that in the experiments with electrodeless discharges, in which there was no possibility of contamination since the vesicles were never broken, no trace was ever found of nitrogen or argon (O'Keefe et al., 1962) although the technique is very sensitive for both gases.

Gentner and his co-workers report that some of the vesicles are found to be void, in the same region where others have pressures of the order of 0.2 atmosphere (2 x 10⁴ N/m²).  It is hard to imagine how the glass could have had sufficient viscosity to maintain a pressure difference of this kind, and at the same time permit the escape of the bubbles required to remove the oxygen (from the ferric-ferrous reduction) and the water.  Under terrestrial gravity, the buoyancy forces are less than total pressure, in the approximate relation of the size of the bubble to 1 m.

The existence of measurable even if small amounts of O₂  is hard to understand if the vesicles were sealed off while hot.  The equilibrium oxygen partial pressure in contact with tektite glass at 1200° K is about 10^-13 atmosphere.

It seems more reasonable to suppose that the vesicles which contain these gases were open to the atmosphere through very slow leaks.  While the tektite was underground, it would be exposed, not to atmospheric air, but to soil air, which is rich in CO₂ and depleted in oxygen.  If the leak was slow enough, the bubble could fill during the 700,000 years that it was underground, and not completely empty during the week-long outgassing employed at Heidelberg.

Muong Nong tektites and soil melting

Barnes and Pitakpaivan (1962) argued that the Muong Nong materials could be regarded as having originated by melting of the local soil.  The soil is at present very different, being mostly Fe₂O₃ and Al₂O₃ like most laterites.  Barnes and Pitakpaivan argue that it may have had a composition like a tektite before laterization.  The argument is precarious, however, because the lateritic composition is a sink toward which soil compositions of all kinds -- even limestones -- trend as a result of strong leaching.

Urey (1963) and Barnes (1971c) regard the Muong Nong tektites as produced in place.  However, the localities are up to 900 km apart, from a point 81 km west of Sakhon Nakhon (Barnes, 1963b) to Bien Hoa, near Saigon (Fontaine, 1966) within Southeast Asia, and up to 2400 km if the Philippine site is counted (Barnes, 1964b).  Since these sites are about 100 times larger than the sizes of comet nuclei, Barnes (1971c) suggested that they are produced by the comae of comets.  A. H. Delsemme, however, points out (Barnes and Barnes, 1973, p. 358) that the density of gases in the comae of comets is so exceedingly low that no mechanical effects of a collision of this portion of the comet with the earth is likely to be felt at the earth's surface.

Oxygen partial pressure

Walter and Doan (1969) calculated by extrapolation of their data that the oxygen partial pressure in tektites would not be far from equilibrium with the earth's atmosphere at 3000° C.  On the other hand, the effect found by Brett (see Chapter 4) is direct evidence that the oxidizing properties of terrestrial sediments are much higher than those of tektite glass. 

Mechanics of terrestrial impact

Arguments based on the mechanics of tektite formation are given in the next chapter, under the heading of arguments against the terrestrial origin, because of the difficulties which they reveal in the hypothesis of terrestrial origin.

Formation of lechatelierite

In his recent book, Rost (1972) offers a most important and fundamental argument against the volcanic origin of tektite glass.  He notes that lechatelierite is usually present in place of quartz or cristobalite in tektites; and that in terrestrial lavas, it is not reported.  On the other hand, he notes that Barnes (1958b) refers to the presence of lechatelierite in Pele's Hair, a volcanic product of Kilauea.  The reference is to Baker(1944), who reports finding it.  His finding has not had recent confirmation.

CONCLUSIONS

Tektites cannot come from beyond the earth-moon system.  They are either terrestrial or lunar.  Arguments against a lunar origin are therefore arguments for a terrestrial origin.

Tektites have more silica in their composition than would be expected from their content of mafic oxides.  This is interpretable either in terms of terrestrial sedimentary processes (which may enhance quartz, as in sandstones) or in terms of lunar magmatic processes (as in sample 12013).

Tektites are more like terrestrial rocks than like most returned lunar samples with respect to lead isotope abundances and rare earth element abundances; they are more like terrestrial rocks than like any returned lunar samples yet reported with respect to oxygen isotope ratios.  It follows that tektites can be lunar only if there are places on the moon which contain more volatiles than the surface areas sampled during Apollo and Luna missions.

Hence it is very improbable that tektites are removed from the lunar surface by meteorite impact, which would be expected to hit the kinds of terrain which are commonest on the moon.  Tektites must be either terrestrial or else the product of lunar volcanism.

A lunar origin for tektites is possible only if the lunar interior contains regions which are richer in volatiles than the areas sampled for the Apollo and Luna missions.  The lead isotope data from Apollo 17 give some color of reality to this idea; but apart from these, the hypothesis of a relatively volatile-rich region in the moon is clearly ad hoc, and therefore somewhat lacking in credibility.