SCIENTIFIC TESTING AT GLOZEL: 1927-2005

The Chemical Analysis of Glass
Glozel glass was first analyzed in the 1920s by Professor Croze, using spectrographic techniques and classic wet chemistry, and by M. Bruet, Vice President of the Geological Society of France. It was analyzed again in the 1990s at the Slowpoke Reactor at Toronto University in Canada by the technique of neutron activation analysis. The new and old analyses are in general agreement. The most recent tests identified seven types of glass, three of them high-potassium glass typical of the French medieval period.

Analyses Made on Glozel Ceramics
I. The Chemical Analysis of Pottery

In 1928 Bruet performed an examination of a Glozel tablet with inscriptions and of a sample of Glozel clay. He found that the same proportions of the same minerals were present in both samples, allowing one to conclude that the tablet was made from Glozel clay.

Archaeomagnetic studies by Barbetti in 1976 showed that the objects could not have been made of reconsituted older ceramic ware. Also in 1976 Zimmerman applied the zircon dating technique to two objects and found that there was no possibililty that they had been artificially irradiated to "age" them.

Tests using differential thermal analysis on fourteen Glozel ceramic artifacts were performed by Vagn Mejdahl in 1980; they established that ten of the fourteen had been fired to at least 500° C.

The recent neutron activation analysis in Toronto of a phallic idol and five inscribed tablets confirms that the objects were made from Glozel clay. Stoneware pottery samples, however, clearly were made somewhere else.

II. The Dating of Ceramics by Thermoluminescence (TL)

The first extensive TL work, carried out in 1974 by Van Mejdahl, Hugh McKerrell, Henri François and Guy Portal, was an authenticity survey of nineteen ceramic artifacts from Glozel. Their joint first paper, authenticating the Glozel ceramics, appeared in the journal Antiquity in 1974 (McKerrell et al 1974).

By 1979 the group had carried out 39 TL datings on 27 artifacts. The dates fell largely into three groups: an early period from about 300 B.C. to 300 A.D.; a medieval period that clustered around the 13th century; and a recent period. Fifteen of the artifacts dated to the early period. These consisted of inscribed tablets, lamps, face urns, vases, and an enigmatic object that Morlet called a bobine.

Eight pieces dated to the medieval period. The original Celtic TL dates of some of these objects might have been reset to the Middle Ages because of nearness to the heat of the fosse ovale, used as a glassmaking kiln. Exposure to heat can reset the TL clock to zero.

Three pieces had recent dates: a lump of clay and a brick with cupules dated to the late 18th century, and a vitrified tablet dated to the 20th century.

In 1983, as part of the new investigation of Glozel by the Minister of Culture, tests were performed at the Oxford TL laboratory on five ceramic samples, again excavated close to the remains of the fosse ovale. The dates ranged from the mid fourth century to the medieval period. In 1985 scientists from Oxford dated five ceramics from the museum. Two vitrified tablets were recent in date; the other three inscribed fragments were medieval.

III. Magnetic Data on Glozel Ceramics

Samples of several tablets were sent to John Shaw at Liverpool University to have their magnetic history analyzed. He was able to show that two of the vitrified tablets as well as a phallic idol, had had an earlier firing, to above 700°C, followed by a later firing to about 400°C. Only the former heating could have effected the vitrification. Although the tablets may originally have been made and fired in the Celtic period, vitrification in a glass kiln would reset the TL clock, erasing any earlier firing.

He also determined that two tablets found in the tombs and dated to the medieval period also had earlier heatings. In this case the first heating would have been in the Celtic period and the second in the medieval period.

Discussion of the Ceramic Analyses
The TL data and magnetic analyses suggest that there are actually two occupation periods at Glozel, not three. The artifacts which seem to be associated with the earliest period are the inscribed tablets, the face urns, the vases, the phallic idols, the bobine, and the lamps.

The medieval TL dates seem to be associated with the glassmakers and are supported by the medieval C-14 dates and by the glass analyses. Hand prints, bricks with cupules, a tile from the floor of the kiln, and some pieces of clay all belong to this period. The recent period might be an artifact of ceramic reheating due to the effects of a large field fire at the time the Champ des Morts was cleared in the 1990s.

Analyses Made on Bone
I: The Chemical Analysis of Bone

In 1928 nine bone objects from Glozel were analyzed by Maheu and Randoin of the Paris Police Laboratory as part of the Bayle report on Glozel. Conversion of their results reveals nitrogen levels between 0.3% and 2.5% for these objects. Although nitrogen decreases with time, there is great variation in the rate of its loss. Modern bone contains about 4%N but Neolithic bone has been shown to contain between 2.9% and 0.1%N.

About the same time Morlet arranged for chemical analyses on eleven Glozel bones to be made at four European laboratories, including those at the Universities of Porto, Olso, and Lyon. Nitrogen levels for these objects ranged between 0.1% and 2.2%.
In 1976 Hugh McKerrell determined the amount of nitrogen in fourteen pieces of bone from Glozel, ten of them undecorated and non-human. The levels were between 0.2% and 0.7%.

In 1997 Hugh McKerrell and Alice Gerard sampled 62 bones from the Glozel museum, twelve human bones and the rest worked or decorated animal bone. Nitrogen levels ranged from 0.5% to 4.2%. One of these artifacts, a fish hook, was chemically analyzed both in 1928 and in 1997. In 1928 it was found to have 2%N; in 1997 2.1%, demonstrating a good correspondence between the older and newer analyses.

II: Carbon-14 Dating

Sometime in the late 1940s, Morlet learned that a new technique - Carbon-14 - could for the first time accurately date old bone. Since bone loses C-14 at a measurable rate once an animal, or human, has died, a measurement of the remaining C-14 can give an age for the bone. After an unsuccessful attempt in 1954 to have bones dated in the United States, in 1957 Morlet sent two human bones to the French radiocarbon laboratory at Saclay for C-14 dating. Both gave recent dates, essentially modern, and the results were never published.

Table 1 below gives the results all of the C-14 dates, in historical order.

In 1975 McKerrell sent an ox tooth that had been found in a Glozel urn to East Kilbride in Scotland for C-14 testing. It dated to AD 30-230, very close to the TL dates being obtained at the time on Glozel ceramics. Than same year a sample consisting of fourteen pieces of bone was also tested at East Kilbride. The date, about 17,000 BC, was never published because the sample appeared to be contaminated with some kind of wax, which could not be completely removed. Fluorine levels found in these bones were exactly the same as in many other Glozel bones recently dated by C-14 to the medieval period, indicating that the very old date was not reliable. The fact that the bones were tested all together as one sample also invalidates the result.

Three C-14 analyses were made at Oxford in 1984. A piece of charcoal dated to 1020-1220 A.D. and a fragment of an ivory ring dated to AD 1400-1490. A portion of a human femur was dated to AD 340-530. The charcoal had been removed from an inscribed tablet already TL dated to AD 1350±125; essentially both dates agree.

In 1995 Alice and Robert Gerard sent two small decorated bone tubes found in Tomb II in 1927 to the University of Arizona for C-14 testing using the Accelerator Mass Spectrometery (AMS) technique, which can determine a date from very small amounts of bone. The two dates were medieval: AD 1250-1300.

Two years later, after McKerrell had begun to work again on Glozel, a new selection of bones to be dated was made. These included a dagger handle decorated with reindeer and alphabetic letters; a bone decorated with a troop of horses and alphabetic symbols; another depicting reindeer confronting each other; part of a human cranium recovered from Tomb I in 1927; and large harpoon made of deer antler with several alphabetic symbols on it. The first three dated to the medieval period, immediately raising suspicion they they were modern fakes made from old bone.

However, examination of the engraved lines on two of these bones with the scanning electron microscope revealed that the cracks in the bone, which appear quickly once the artifact emerges from burial in the moist Glozel soil, cross the engraved lines on the piece. Therefore the engraving appears to have been made when the bone was fresh, in the 13th century. The cranium, like the earlier human bones dated at Saclay, dated to AD 1650-1950 and the harpoon to AD 1520-1650. McKerrell also sent a piece of carbon from a vitrified tablet (984.2.006) to Arizona for dating. The date obtained, more than 46,000 years BP, suggests that the carbon was actually coal or coke, neither of which were available to 13th century glassmakers. It is possible that the coke came from a forge used in the 1920s to separate the vitrified tablet from an idol which was stuck to it by melted glass. Early pictures show them attached; today they are separate.

In 2000 René Germain arranged to have three pieces of human bone from Glozel dated at the Arizona laboratory. GF745, part of a the same cranium as the previously dated GF743, dated to 1850-1955 AD. GF755, a mandible, dated to AD 1460-1640 and GF737, a fragment of a left femur, dated to AD 1440-1525 (personal communication).

This short summary of the Glozel scientific analyses raises many questions. Some are answered in the complete text titled Scientific Analyses at Glozel, which is available on this site as a pdf file. Others can only be answered by more research.