Tektites and Their Origin - Chapter Two

TEKTITE DISTRIBUTION:  THE STREWN FIELDS AND THEIR GEOLOGIC RELATIONS

The distribution of tektites over the surface of the earth is not random.  In this they differ sharply from the usual meteorites, as can be seen from a comparison of Figs. 3 and 4.  The clusterings of meteorites are only apparent, and are associated with areas of high industrial development.  Thus meteorite discoveries, both falls and finds, are commonest in Europe and North America, while tektites are commonest in Southeast Asia and the Philippines.  The meteorite finds are clearly controlled by the intensity with which the search for meteorites is conducted; on the other hand the distribution of known tektite finds is the revelation of a geographic pattern which is imposed by nature.  Despite the intense geologic effort that has gone on for centuries in England and Germany, not a single tektite has been found in either country, yet a single cubic meter in the Philippines yielded over 100 (Chao, 1964a).

In place of a random distribution of tektites, what is found is a distribution into what are called strewn fields.  Each strewn field corresponds to a single event, which most workers would say was the fall of a large number of tektites.  The neutral word strewn field was adopted at a time when a considerable number of workers considered that tektites might be produced in the localities where they were found, e.g. by volcanism, by artificial means, or by desiccation of a silica gel.

There are four (or possibly only three) generally recognized strewn fields, plus two (or possibly any number from zero to four) minor distributions of tektite-like material which may constitute strewn fields.  The fields are shown on Fig. 3.  The three major fields are the Australasian field, formed about -0.7 million years, the Czechoslovakian field, formed about -15 million years, and the North American field, formed about -35 million years.  The Ivory Coast strewn field, formed about -0.9 to -1.0 million years, is regarded by D.R. Chapman as perhaps an extension of the Australasian strewn field; we shall here follow the usual practice of regarding it as distinct, but Chapman's doubts are to be kept in mind.  Similarly we here regard the Australian tektites (australites) as part of the Australasian strewn field in spite of the serious and weighty objections brought forward by all Australian geologists who have actually examined the occurrences.  The minor fields are the Darwin-Macedon glass field, formed about -0.7 million years, which includes specimens found chiefly in Tasmania, but with a few pieces 500 km away in Australia; the Aouelloul glass, formed at a date not firmly established, and found only near a small impact crater in Mauritania; and the Libyan Desert glass, found in the Sand Sea of Egypt.  The Darwin glass is considered by many to be part of the Australasian strewn field; we shall here follow that idea, for reasons to be given below.  The Aouelloul glass is very similar to the Darwin glass; it is not usually counted among the tektites,but should probably be counted with them because it presents the same problems, particularly in the field of glass technology.  Similarly the Libyan Desert glass is regarded by many students as a simple result of impact on the desert sand; but this view again encounters great difficulties from the point of view of glass technology and aerodynamics; hence it is logical to discuss this glass with the tektites.

THE AUSTRALASIAN STREWN FIELD

This enormous pattern (Fig. 5), covering about a tenth of the total surface of the earth, has been discovered piece by piece.  (For the separate portions, see Figs. 6-10.) [editor's note:  seems to me that should have read Figs. 6-11 - possibly a typo in original. - LH]  For each portion of the strewn field a special name is employed (Table I).

Have we found the true limits of this field?  Despite earlier suggestions by David et al. (1927) that all tektites result from the passage of a group of natural satellites of the earth over a single great-circle path, it is now clear that the dates of the moldavites and the North American tektites exclude them from the Australasian strewn field.  Tektites have been sought in vain in New Zealand (Eiby, 1959); the boundary of the australite strewn field in Australia is very sharp (Fenner, 1940a); tektites have been mentioned but never brought in from New Guinea (Von Koenigswald, 1960b); the alleged tektite from Timor proved to be an obsidian (Wichmann, 1882); the Sakado glass from Japan (Baker, 1959b) appears to be a terrestrial crystalline volcanic rock with a film of glass; in Africa a number of specimens have been brought in as tektites and later identified as indochinites (Preuss, 1969; also an unpublished study of a tektite reportedly from Nigeria) or as terrestrial igneous rocks (Saul and Cassidy, 1970).  The best indication of the limits of the field is perhaps the negative results of oceanic cores outside the field (Glass, 1972a; see Fig. 5).

Unity of the Australasian strewn field

Internally, there are strong reasons for regarding the Australasian strewn field as a unit.  The external shapes (including the sculpture of pits, etc.) of the tektites form a continuous sequence, with blocky Muong Nong-type tektites in the north, decorated splash-form tektites in the center, and smooth flanged buttons in the south and east (Von Koenigswald, 1967).  (For the meaning of these terms, see Chapter 3.) 

The Australasian strewn field consists of streaks of tektites of similar composition extending for thousands of miles, as was shown by Chapman (1971; see Fig. 11) [editor's note:  seems to me that is a typo, and should read Fig. 10 - LH].  Chapman finds essential identity in composition between tektites at points thousands of miles apart (e.g. West Australia and the Manila area, or Kuchenari, Thailand, and Port Bayard, South China).  These streaks of constant composition tend to run perpendicular to the lines of constant morphology described by Von Koenigswald (1967), as would be expected if the streaks represented the paths of groups of bodies and the morphology recorded some ballistic parameter, such as velocity, angle of entry, or the like.  There is no correlation with the underlying rock (Fenner, 1940b).

The results of Chapman are so remarkable that it is well to note that they are foreshadowed by Heide's finding (1936b) that there is a central streak of nickel-rich tektites going from Indochina to Java, flanked on either side by streaks of low nickel content (Philippines to Australia, and Thailand); they are similarly foreshadowed by the finding of Schnetzler and Pinson (1963) that indochinites differ systematically from philippinites, especially in calcium content.  Again, there is a note of S.R. Taylor (1964) on an isolated patch of high-nickel australites which, in the light of Chapman's work, is seen as a part of one of his streaks.  Again Tatlock (1965) noted some chemical correlations between tektites of Western Australia at Kalgoorlie and the tektites from the vicinity of Manila.

In support of the chemical studies, Chapman et al. (1964) made plots of the density versus frequency, which he called population polygons of specific gravity.  These plots showed the same relations between regions that were found from the chemical composition.  That is, regions with similar chemical composition also had similar population polygons of specific gravity.  The orderly pattern shown by all these studies is strong evidence for the unity of the Australasian strewn field.

If the glass found at Macedon, Victoria, Australia (Baker and Gaskin, 1946), belongs with the Darwin glass, as appears from its composition (Chapman et al., 1967), then this may represent a streak of high-silica glass very roughly parallel to the overall structure of the Australasian strewn field in this region.  The Darwin glass itself is distributed along a narrow north-south streak in Tasmania (see Fig. 11) [editor's note:  I think this reference should be to Figure 10 - LH].  Recently an apparent crater has been reported in this region, with abundant glass near it.

Another approach to the problem of the unity of this field comes from studies of the dating.  Potassium-argon methods and fission-track methods (see Chapter 7), which measure the time since the last strong heating, give accordant dates of -600,000 years to -800,000 years for all types of tektites in the strewn field, except the high-sodium tektites.  For these a date of -3.7 million years is found by fission-track methods. The Darwin glass appears, on this basis also (Gentner et al., 1972), to be part of the strewn field.

Despite the above, Australian geologists are agreed that the australites arrived at the earth at a date nearer -14,000, rather than -700,000 (Fenner, 1938, 1949; Baker, 1962; Gill, 1965; 1970a; Lovering et al., 1972).  The papers of Gill and of Lovering et al. supply clear evidence that tektites are found on top of recent Australian soils whose ages, as given by carbon dating, are less than 20,000 years.  The evidence is strong that they did not reach this position by reworking from older sediments at a higher elevation. The remarkable state of preservation of the fine markings on some of the australites indicates clearly that they have moved only very small distances at most.  For example, a Czechoslovakian study (see below,  p. 29) [LHancock editorial note - O'Keefe's "p. 29" refers to the section below entitled, "The Moldavite Strewn Field"] shows that stream erosion will reduce glass objects of roughly tektitic character to about one-ninetieth of the original mass at a distance of 40 km downstream.  Near Lake Torrens, Lovering et al. (1972) found well-preserved australites at a place where, according to the accepted chronology, the soils formed at -700,000 years are deeply buried.  From their maps, it is seen that the nearest surfaces which are older than -700,000 years are 15-20 km away.

Gill (1965) surveyed a single square chain near Port Campbell, and excavated it carefully.  He found fourteen australites, all in the layer just above hardpan; the latter was dated by some carbonized stems at -5430 years.  The layers with the australites were dated at -3750 or later.

Hodge-Smith (1939) remarks that on the gibber country plains (plains covered by wind-facetted pebbles) all stones show more or less uniform polish and weathering, except australites; some of them show no weathering while others have only traces of their original form.

One obvious solution to this puzzle can be ruled out.  The tektites could not have floated around in space for a few hundred thousand years before falling to earth; if they had, there would be clear chemical and physical evidence of attack by primary cosmic rays.  This has been carefully sought for (see Chapter 7) and not found.  Moreover the flanges of the flanged australites, which appear to have formed during the descent through the atmosphere (Chapter 3), have been dated by fission-track methods at -700,000 years, identical with that of the core (Storzer and Wagner, 1969). 

Thus the age discrepancy remains as an interesting and significant puzzle.  It appears that we must reject the very recent dates for the Australian tektites:  something must be wrong, conceivably the dating of the hardpan.  In this book I shall treat the australites as part of the Australasian strewn field; but it is to be kept in mind that the problem is not solved.

The relation of the microtektites to the larger tektites of the Australasian strewn field is regarded as doubtful by some workers, particularly in Australia (Baker, 1968b).  However, the demonstration by Frey et al. (1970) that microtektites have the same trace elements as the larger tektites has given strong support to the supposed relation.  The microtektites have been dated by study of the magnetic behavior of the sediments in which they are found.  The sediments record the reversals of the direction of the field.  The microtektites are found at the Matuyama-Brunhes reversal (Glass and Heezen, 1967) which has been dated by comparison with land lava flows at about -700,000 years.  The date is the same throughout the field; note that microtektites have been found both north and south of Australia.  The bottle-green microtektites are to be included in the strewn field also, according to the work of Glass (1972b) to be discussed further in Chapter 6.  Similarly the chemical work of Barnes (1964c) shows that the Muong Nong materials, despite their different appearance and internal structure, are chemically identical with the other indochinites and must be regarded as part of the strewn field.

Glass (1970a) found a sequence of compositions within a single australite which parallels closely the sequence of compositions of microtektites from the Australian Basin.  Microtektites from other parts of the Australasian strewn field are not nearly as close.  This strengthens further the identification of microtektites with ordinary tektites; but at the same time it fixes the date of arrival of the australites as the same as the microtektites, namely -700,000 years.

THE IVORY COAST STREWN FIELD

In Fig. 12 is shown the Ivory Coast strewn field (Lacroix, 1934b, 1935a), together with the Bosumtwi crater in Ghana, from which many investigators think that these tektites are derived.  Also shown are the locations of two oceanic cores in which Glass (1968, 1972b) found microtektites.  The chemical compositions of the microtektites, particularly at the high-silica end, clearly relate them to the ordinary tektites of this strewn field.  The total extent of the field is not well known; much of the land area is heavily wooded, and the tektites are not found at the surface but at depths of 5-6 m in alluvial deposits.  A careful guide to the strewn field was prepared by Saul (1964).

Only a few hundred Ivory Coast tektites are in scientific collections; all of these are splash-form tektites; neither Muong Nong tektites, nor flanged buttons and related forms have been found among the Ivory Coast tektites.

It is important to see that the discovery of the microtektites means that the center of the known field is 900 km from Bosumtwi, rather than 300, as earlier believed; and the field extends to a distance of 1600 km from Bosumtwi.  Moreover, the angle subtended at Bosumtwi is about 49°; thus one can no longer say that the tektites are found on a line radiating from Bosumtwi.

According to standard fallout tables (Glasstone, 1962), a particle with a diameter of 200 μm is expected to fall to the ground in less than 2 hours.  Since the microtektites include particles as large as 1 mm, it is clear that even very strong winds cannot have significantly altered the form of the strewn field.

THE MOLDAVITE STREWN FIELD

The moldavites cover two small patches in southern Czechoslovakia (see Fig. 13), one in the territory called Bohemia, the other in Moravia.  The field has been described in detail by Vorob'yev (1964).  The name comes from the German name Moldau for the Vltava River in Bohemia.  The Czechs call these tektites vltavines.

Because of its occurrence in Central Europe, the moldavite strewn field has been extensively studied although the mass of glass involved in the ordinary tektites is only about 3000 tons (Bouška and Rost, 1968).  Storzer and Gentner (1970) have interpreted some bentonite particles found in the Bavarian Molasse deposits (late orogenic sediments) as micromoldavites chemically altered by contact with water.  The data are still preliminary.  About 200 km from this field, in a direction south of west is the Ries Kessel, a large impact crater of late Miocene age which is regarded as the source of the moldavites by some investigators.

A valuable study of the process of destruction of glass during transportation by water was made by a group of Czechoslovakian students (Anonymous, 1971) who studied the rate of loss of weight for pieces of artificial glass in streams as a function of the distance downstream from the factory which made them.  The students found that the glass lost about 99% of its mass in a distance of 40 km.  This study suggests that the moldavites cannot have been moved very great distances – probably not more than 10 km.  Bouška et al. (1968) reach similar conclusions.  It follows that the moldavites must have fallen near the locations in which they are found.

The Bohemian moldavites are slightly different in chemical composition from the Moravian moldavites.  Each of the two small strewn fields is elliptical, with the major axis in the NW-SE direction.  In each of the subfields, there is a distinct gradation in size, so that the specimens from the northwest end of the subfield average about three times as heavy as those from the southeast end (Šimon, 1963; Bouška et al., 1968).  These differences are much larger than the differences in average weight between the two subfields.  Qualitatively, the subfields are like the fields produced by meteorite falls (Nininger, 1952), although the gradient in size is not as steep as in meteorites, and the field is relatively long.  If this is the explanation then the moldavites entered the atmosphere at an angle over 90° to the direction from the Ries crater.  Bouška et al. (1973) note this possibility.  They seek, however, to explain it on the ground that the moldavites have been reduced in size by stream erosion.  Their explanation seems somewhat artificial since the Vltava (in Bohemia) flows north, while the Jihlava (in Moravia) flows southeast, but the heavier tektites are at the northwest in both fields.

In further support of a NW-SE orientation of the strewn field is the finding of two moldavites in the northern suburbs of Prague (Rost, 1972, p. 93) and perhaps even the three flaked moldavites found at Willendorf (Bayer, 1918).

The geologic age of the strata in which they are found is late Miocene.  Janoschek (1937) stated that near Dukovany he had found layers of tektite-bearing gravel interfingering with marine deposits which had fossils of Oncophora; these would give a date of late Helvetian.  Bouška (1964) revisited the site, and found that the moldavite-bearing gravels overlie the Oncophora layers.  Žebera (1968) studied some ancient lake bottoms, where he found occasional pairs of tektites, consisting of fragments of a single original piece, the two fragments being only a hundred meters or so apart.  These, he feels, cannot have resulted from redeposition, which would have separated the fragments much further; and hence these deposits are where the tektites first fell.  They are found in Vrabce clays, above the Mydlovary formation, which is of lower Tortonian age.  The Tortonian is Upper Miocene, and just above the Helvetian.  The catastrophe which formed the Ries is also of Tortonian age (Preuss, 1964) and hence cannot at present be distinguished in time from the moldavites.

K-Ar ages for the moldavites and the Ries also agree (Gentner et al., 1963; see Chapter 7).  Thus the geographic data point away from the Ries as the source of the moldavites, while the chronological data point toward the Ries.

THE NORTH AMERICAN STREWN FIELD

The core of our knowledge of the North American strewn field comes from the bediasites (see Fig. 14), found in Texas in a narrow strip of land paralleling the Gulf Coast, and about 150 km inland. The bediasites are found close to the outcrop of the Jackson formation, which is uppermost Eocene.  Tektites from the southern part of the bediasite area, near Muldoon, are often lighter in color (Barnes, 1951). Chao (1963) notes that they are associated with a lag gravel characterized by a reddish chert.

Tektites of related age and composition in Georgia were first reported by Bruce (1959), chiefly in Dodge County; one was also found in Washington County (Pickering and Allen, 1968).  A large number were found by Howard (1968).  They are found, as King notes (1962), in a surface gravel overlying the Hawthorne mottled clays.  The difficulties in explaining this problem are noted by Furcron (1961).  King (1962) attributes the gravel to a "Pleistocene-Pliocene wash" covering the whole area.  In fact, the tektites are typically found, not on stream terraces, but on the divides.  Personal examination convinced me that the gravel is a lag gravel; one sees everywhere pebbles on the top of small earth pillars (demoiselles) of Hawthorne clay.  In the ditches one can see how the gravel is being produced by the washing of the clay.  The sharp contact noted by King (1962) is an isolated example of a stream deposit; at most tektite sites the contact between the gravel and the clay is gradational and is only a decimeter or so below the surface.  The Georgia tektites seemed to me to be weathering out of the Hawthorne.  Most of those in Dodge County seem to come from points between the 300- and the 325-foot contour on the 1:250,000 map.

The Hawthorne is generally assigned to the Miocene; but it shows no fossils within about 100 km of Dodge County.  In this county it overlies a limestone which is called Oligocene.  However, the well logs, which have been studied and compiled by Herrick (1961), systematically note the presence of Middle Eocene foraminifera in this limestone.  Although the well logs note the possibility that these foraminifera may have been reworked, S.M. Herrick (personal conversation, 1973) remarks that they do not appear reworked.  It is thus not inconceivable that the lower Hawthorne is of late Eocene age, or Oligocene.

The single Massachusetts specimen is from Gay Head, on Martha's Vineyard (Kaye et al., 1961).  It was found in a gully, clearly displaced.  It was 8 m below the top of sands dated as Raritan (early Late Cretaceous).  Above them is the Aquinnah conglomerate, itself a reworking of Cretaceous and Miocene deposits; above that is Middle Pleistocene and Holocene sand.

Microtektites apparently related to the North American strewn field have been reported by Donnelly and Chao (1972) and Glass et al. (1972a).  The cores are in the Caribbean, near the island of Curaçao.  In addition, there is a single tektite from Cuba (Garlick et al., 1971) which seems to be compositionally related to the North American strewn field.  It was doubted at first; but the same chemical peculiarities which raised the doubts appear in some North American microtektites; others have compositions whose relation to the bediasites is close.  The date of the microtektites is again late Eocene, as judged from the stratigraphic data in the cores.

From the geological standpoint it is interesting to note that the North American strewn field can be reliably placed in the late Eocene, since the K-Ar and fission-track ages are about 35 million years at most, whereas Kulp (1961) places the top of the Eocene at -36 million years.

THE THREE MINOR FIELDS

Libyan Desert glass

Clayton and Spencer (1934), acting on reports from the Survey of Egypt, visited this difficult location (Fig. 15) and collected about 50 kg of this glass.  Spencer (1939) returned to the area and estimated that the field extends from 25°2'N to 26°13'N, and from 25°24'E to 25°55'E, an area 130 km north and south by 53 km east and west.  Isolated pieces, some at least certainly transported, were found at distances of some hundreds of kilometers.  The material probably exceeds the moldavite strewn field both in quantity and in area.  The age is 28.5 million years, by fission-track dating (Gentner et al., 1970a).  Its relation to other tektites is open to some question, because it is nearly pure SiO₂.  The reasons for grouping it with the tektites are discussed in later chapters; the principal reason is just the difficulty of explaining how large chunks (up to 7 kg) of this very viscous glass could be produced and freed of volatiles by natural processes on the earth.

The material is found between dunes on the flat desert floor in a reddish soil containing much calcium and magnesium carbonate, gypsum, ferric oxide, and clayey materials.  The soil overlies the so-called Nubian sandstone [according to Pomeyrol (1968) the term is so widely applied (Libya to Sinai, Cretaceous to Tertiary) as to be meaningless].  The dunes, which are very sharply set off from the desert floor, consist of yellow sand.  On their tops are found fulgurites (glass tubes formed by lightning), which are entirely different in structure from the Libyan Desert glass, being thin-walled and full of bubbles.

A crater exists in the vicinity, at 22°20'N and 25°30'E.  Meteorites were sought in the area of the Libyan Desert glass by Spencer, without success.  The crater and the strewn field were revisited recently by Barnes; the results are not yet published.

Aouelloul crater glass

A small quantity of glass, resembling Darwin glass, is found in the immediate vicinity of the 250 m crater of Aouelloul, in the Mauritanian Adrar (Sahara), at 20°15'N, 12°41'W, in level sandstone of early Ordovician age, far from any volcanic site.  The crater is generally regarded as an impact crater (Campbell-Smith and Hey, 1952a,b).  The question has been raised by Campbell-Smith and Hey whether we are to regard the crater as excavated by a block of glass, or to regard the glass as formed by meteorite impact on the local (Zli) sandstone.  It is classed with tektites because of its similarity to Darwin glass, and because of the difficulties of glass formation in the very brief moments of a meteorite impact.

The glass is found around the crater, especially on the east side.  It extends at least 1 km eastward; extensions in other directions have not been noted.

The K-Ar age is 18.6 million years (Gentner et al., 1968); the fission-track age is about 0.3 million years (Fleischer et al., 1965b).  The freshness of the crater suggests that it was formed at -0.3 million years; the higher K-Ar age would then mean that argon was preserved through the crater-forming event.  The fission-track age may also be too low; Cressy et al. (1972) refer to a determination of 3.5 million years by Storzer, who corrected for track fading.

The Ordovician locally comprises two layers, O₂ (Oujeft sandstone) and O₃, the next higher layer (Zli sandstone).  The O₃ is distinguished from the O₂ by the presence of vertical fossil burrows (tigillites; Monod and Pourquié, 1951).

Amerikanites

There is a class of glass bodies, often spheroidal in shape, having markings resembling those on splash-form tektites, found in Peru,  Ecuador, and Colombia (Stützer, 1926; Codazzi, 1929; Martin and De Sitter-Koomans, 1956) which are called macusanites or amerikanites.  They are connected by chemical composition with local lava flows; in some cases they seem to represent sediments which have been melted and then erupted.  They are not accepted generally as tektites; however, if it should turn out that tektites are the product of lunar volcanoes, it is possible that the amerikanites are analogous terrestrial objects.