Levels of uranium oxide in modern bone are practically nil, but in ancient buried bone these may rise as high as 1000 ppm [parts per million] depending on the concentration of uranium oxide in percolating water. Aitken gives the range in fossil bone as lying between 1 and 1000 ppm. Trace amounts of fluorine are present in modern bone, ranging from 0.01 to 0.1 per cent in human bone, from 0.024 to 0.07 per cent in adult dentine of tooth, and between 0.02 to 0.1 percent in Red Deer bone. Thus the maximum ever in modern specimens of tooth and bone in man and deer is 0.1 percent.62
Nitrogen
This test works the opposite way round from the other two. Unlike fluorine and uranium, which can only begin to accumulate from the surrounding soil and its percolating water after the death and burial of the organism, nitrogen is accumulated in bones and teeth etc. only during life and then begins to dissipate. Thus after death the general rule is that ‘nitrogen decreases with increasing bone age’:63
Bone and teeth contain a certain percentage of nitrogen, averaging 3.4 per cent in teeth and 4 to 5 per cent in bone. Following death and burial organic remains lose their nitrogen with time, once the requirements of its breakdown are available. These include the absence of glacial conditions, an alkaline medium, absence of surrounding clay, and the presence of a specific bacterium, the Clostridium histolyticum …64
Example: if human teeth and deer bones and/or antlers are excavated from the same horizon and if the teeth prove on testing to contain much higher levels of nitrogen than the deer remains, then the implications would be that the teeth must be much younger than the deer remains and are thus intrusive to the horizon. The provisos, however, are many: if the environment lacks the bacterium necessary for nitrogen breakdown, or is glaciated, or surrounded by clay, then nitrogen is retained in any buried teeth and bones and the depletion of nitrogen becomes less useful as a test of relative antiquity.
‘In effect, therefore,’ Mifsud concludes,
a low nitrogen is useful to indicate antiquity, whereas a high nitrogen is not significant unless it is associated with a low fluorine and uranium oxide, which will definitely indicate a recent specimen. Conversely, the presence of a low fluorine and uranium oxide is not significant in the presence of a low nitrogen, for there are factors which impede fluorine and uranium oxide uptake, particularly in limestone caves of which Ghar Dalam is one. On the other hand a high fluorine and uranium oxide is significant in reflecting antiquity.65
Politics and ambition
Mifsud’s view is that the results of the 1952 FUN tests on the Ghar Dalam teeth got caught up in a number of matters incidental to the proper concerns of archaeology that made it expedient for them either to be ignored, or better still discredited, as evidence of a Palaeolithic human presence on Malta. Of these the two most important were local politics on the one hand and the academic ambitions of the late Professor J. D. Evans on the other.
As to politics, Malta in the early 1950s was pursuing an integrationist policy with Britain. Absurd as it seems now there was embarrassment in official circles that the taurodont teeth might prove the modern Maltese to be directly descended from primitive Neanderthal ancestors (although, as we’ve seen, taurodontic teeth on their own, even specimens as large as the Ghar Dalam molars, do not necessarily prove that the original owners of those teeth were Neanderthals, since the condition still exists to varying degrees in modern, non-Neanderthal humans today).
As to the second matter, Mifsud notes that J. D. Evans had graduated from Cambridge in 1949 and that in the early 1950s he was ‘in desperate need of a PhD’.66 The thesis that the future Professor of Prehistoric Archaeology at the University of London chose to develop, influenced by the Italian archaeologist Barnarbo Brea, was that the very first human inhabitants of the previously unpeopled Malta had been immigrants from the Neolithic Stentinello culture of Sicily – a theory that is still part of the conventional academic wisdom about Malta today. In pursuing this thesis, Mifsud suggests, it was not convenient to the young Evans to have to deal with the evidence of the Ghar Dalam teeth that suggested a prior, Palaeolithic, human presence in Malta.
This, then, either as a conscious or unconscious motive, could explain why Evans was so vehement in his attacks on the antiquity of the taurodonts and so economical with the truth in his published statements about them. He wanted them out of the way – permanently – of his own theory about Malta’s first inhabitants.
A tale of two museums
To get to see the records of the 1952 tests Anton Mifsud at first expected that he would need to travel no further than the distance from his own home to the Valletta headquarters of the National Museum – on behalf of which the Natural History Museum in London had carried out the tests in 1952.
This turned out to be rather a naive expectation. But what Mifsud did discover in Valletta was that:
On 3 March 1952 Dr J. G. Baldacchino registered the sampling of the taurodont molar discovered by Despott in 1917. Other remains from Ghar Dalam cave included another tooth which was picked up by Caton-Thompson in 1924, a taurodont molar elevated by Baldacchino in 1936, and one sample each of a hippo molar and deer longbone. There is no record in the Museum of Archaeology Reports [for 1952/3] of these tests being carried out.67
At this point Mifsud flew to London, where he found, to his relief that:
The Green Book at the Museum of Natural History is still available and contains the original readings of the entire repertoire of tests carried out between 1952 and 1968/9 on the ‘Malta Samples’. Two teeth from the Hypogeum were also included.
The five human teeth submitted to the Natural History museum were therefore Caton-Thompson’s (Ma.1), Despott’s (Ma.2), Baldacchino’s (Ma.7) and two molars from the Hypogeum (Ma.5 and 6).
The results in the Green Book were not what Mifsud expected. The fluorine test for Despott’s molar (Ma.2) gave the highest results of all the samples tested, including the Pleistocene deer and hippo samples. This result did not jibe with the official position that the ‘careful chemical tests’ had proven the tooth to be Neolithic. The nitrogen result for Despott’s molar was 1.85 per cent. Were it not for the fluorine readings, this result would have been compatible with the official position. But as it stands, the disparity revealed in the Green Book only suggests that either the fluorine tests or the nitrogen tests – or both – were unreliable. The official position should therefore have been that the results for this tooth were internally inconsistent and hence ambiguous.
Baldacchino’s molar gave very similar nitrogen results to the two hippo molars from Ghar Dalam (0.44 per cent compared with 0.4 per cent for the hippos), clearly suggesting contemporaneity between the human samples and the Pleistocene animal samples in the Cervus Layer. Again, this result is incompatible with the official claim that these chemical tests proved the human teeth in the Cervus Layer to be Neolithic intrusions.
The tooth Ma. I was discovered in the mid-1920s by Gertrude Caton-Thompson, the British archaeologist cited earlier whose views were misinterpreted by Evans. According to Caton-Thompson’s notes, it was found in an ‘unstratified layer’ in the company of hippo, horse, deer, thirty potsherds and the end of a flint blade. This tooth yielded fluorine results (0.2 and 0.3) equivalent to those of the Pleistocene deer samples (0.25 and 0.3). It also yielded similar nitrogen results to Baldacchino’s molar, with two different tests yielding results of 0.39 per cent and 0.79 per cent. That the same tooth yielded such different results highlights the unreliability of the nitrogen testing in a similar manner to the fluorine-nitrogen inconsistency of Ma.2. Nevertheless, these nitrogen results overlap with the readings for the Pleistocene hippo samples (0.4 per cent), and further raise the possibility of a Pleistocene date. This is especially interesting because the tooth bore no signs of taurodontism, and might therefore be taken as evidence of a Palaeolithic presence of humans on Malta who had normal, non-taurodontic teeth.
It is worth noting here that another non-taurodontic tooth was discovered in Ghar Dalam in the
1920s, this time by George Sinclair, a civil engineer with the British Admiralty. The tooth was buried almost a metre deeper than Despott’s 1917 molar,68 and it is unfortunate that it was not also submitted for chemical testing.
Ma. 6, a human tooth from the Hypogeum, gave a nitrogen reading of nil. If we were to base everything on Oakley’s nitrogen test, we would have to conclude that the owner of this tooth was alive way back into the Palaeolithic. However, it must be pointed out that Mifsud later managed to get this tooth carbon-dated through the Natural History Museum. As noted in chapter 17, the carbon-date put the tooth into the Late Tarxien phase, around 2200 BC. Along with the inconsistencies noted for Ma.1 and Ma.2, this further highlights just how unreliable Oakley’s nitrogen dating technique can be.
The 1964 Report: erasing the Palaeolithic peril
No official record of the chemical testing was published until the Museum of Archaeology’s 1964 report – a decade after the nitrogen and fluorine results had been achieved. During this lengthy hiatus only a very small number of people knew that the 1950s tests had ever been carried out at all and even fewer could have been aware of their results.
Such a delay in the publication of important modern dating evidence confirming a Palaeolithic human presence in Malta is plainly odd in itself. But that the gist of the evidence should subsequently have been misrepresented by the omission of crucial data when publication finally came about is far more extraordinary. Moreover, Mifsud believes the timing of publication in 1964 was not accidental. In that year there was already a dating furore in the air following the discovery that C-14 underestimates the age of materials that are more than about 3000 years old – and that there is a progressively larger underestimation the older the sample is. By 1964 this ‘built-in’ error had been accurately calibrated millennium-by-millennium by means of ‘dendochronology’ (comparison with the annual ring counts of very ancient species of trees). The implications of the new ‘calibrated’ dates for Malta were that the entire Temple Period suddenly had to be shifted a full millennium back in time. For example, before 1964 Gigantija was thought to be no older than 2500 BC; after 1964 and ‘the tree-ring revolution’ the date was pushed back to the presently accepted figure of 3600 BC.69
It is interesting to note that Evans was very slow to accept the implications of dendochronology for his carefully worked-out sequence for the Temple Period (the beginning of which he had hitherto set at 2500 BC) – and even as late as 1971 he was still refusing to let go entirely of his pre-calibration scheme.70 But the tree-ring revolution was an irresistible force, like a rising tide – with implications for radiocarbon-dates all around the world – and even Evans in the role of King Canute could not hold back the waves.
The FUN tests in the 1950s were quite a different matter, done behind closed doors, strictly between the Natural History Museum in London and the National Museum in Malta. By 1964 the extremely annoying and inconvenient results of these tests had been withheld from the public for ten years with no one else any the wiser. Accordingly, there was no basis for protest when the National Museum published an abridged and unfortunately highly misleading version of the results in its 1964 Scientific Report. Whether by accident or by design, the net effect was that only information which supported the Evans paradigm was available on the public record.
The relevant passage from the 1964 Scientific Report itself reads as follows:
Considerable help has been received from foreign experts in the analysis of Maltese material of various sorts.
Dr K. P. Oakley of the British Museum, Natural History, analysed a number of bone samples for their collagen content, expressed as a percentage of nitrogen. The figures obtained were – hippopotamus bone, nil; deer antler 0.13 per cent; normal human tooth 0.7 per cent; taurodont human tooth (these four all from Ghar Dalam) 1.85 per cent … This proves conclusively that the taurodont tooth is later than the material from the other prehistoric sites, and so cannot possibly be of Neanderthal man.71
This statement contains paradoxes which seem all the more bizarre because they are left unacknowledged.
Firstly, the report only makes a conclusion about one human taurodont -Despott’s molar classified as Ma.2 – and yet, as we’ve seen, Trump has used the results of these ‘careful chemical tests’ to draw a conclusion about two different Ghar Dalam taurodonts.
Secondly, why were the obvious inconsistencies in the data ignored? Caton-Thompson’s molar is reported as having a nitrogen reading of 0.7 per cent, when in fact, according to the Green Book, it yielded the two very different results of 0.39 per cent and 0.79 per cent. More importantly, why did the report provide only the nitrogen reading for Despott’s molar, and ignore the contradictory fluorine result? The nitrogen test had already proved itself capable of producing variable and hence unreliable results, so why was it given automatic and exclusive preference over the fluorine results?
Thirdly, if the nitrogen content of 1.85 per cent is supposed to be appropriate to the Neolithic, as we are effectively being told here, then does it not follow that the reported reading of 0.7 per cent from the normal human tooth from Ghar Dalam is indicative of a much older, pre-Neolithic date? And what does it say about Baldacchino’s molar, which gave a nitrogen percentage of 0.44 per cent?
The results published in the 1964 Report misrepresent the actual set of results recorded in the Green Book in London. Had the complete set of results been included, or properly summarized, then it would have been clear that the results were to a large extent ambiguous, but also suggestive of a Palaeolithic human presence on Malta.
This misrepresentation of the actual results of the 1950s tests has subsequently had a pivotal effect on public perceptions of Maltese prehistory and on what university archaeology departments do and do not see as valid and worthwhile research on Malta. Unedited, the results from the chemical tests might have inspired a new generation of archaeologists to break away from J. D. Evans’ 1950s ‘Neolithic’ paradigm and pay more attention to the possibility of much older relics around the Maltese islands – even underwater. But in the distorted form in which the results finally reached the public in 1964 (a mere decade after the tests had been carried out) there could be no danger that they would do any such thing.
Let’s note in passing that seven years after the tests – and five before the misrepresented test results were first put on the public record in the Museum of Archaeology’s 1964 Report – J. D. Evans had begun to talk as though conclusive results were already on the record. Here are three characteristic passages from his Malta (1959):
There are as yet no trustworthy traces of the presence of man in Malta before the Neolithic period …
We have no reliable evidence that any of them [Palaeolithic humans] made their homes in Malta …
We have no reason to suppose that Palaeolithic man ever set foot on Malta.72
‘The logic of Evans’ conclusions’, comments Mifsud, ‘was founded on false premises and a significant iota of misrepresentation … The weight of authority established his hypothesis as semi-dogma; the consequence was bad history.’73
The uranium control
We’ve seen that the fluorine and nitrogen results for Despott’s molar (Ma.2) contradicted each other. The former suggested a Palaeolithic date, whereas the latter – the one which was published – suggested a Neolithic date. But a third test was later carried out on Despott’s molar that had not been carried out on the other two teeth in 1952. This was the uranium oxide assay – a more sophisticated procedure which was not yet fully established in 1952 and which was only applied to Despott’s molar in 1968. This later test took place, Mifsud has discovered, at the specific request of Kenneth Oakley – who also asked that it be carried out on Baldacchino’s (1936) taurodont molar at the same time.74
The uranium oxide result for Despott’s molar supported the flourine result and embarrassingly contradicted the high nitrogen result that had been published in the Museum’s Scientific Report in 1964 as proof that the tooth was
Neolithic. The result of the uranium assay was 13 ppm, compared with 0.1 ppm or less in living bone and levels of between 4 and 12 ppm in various Pleistocene hippo and deer samples from Ghar Dalam. Ghar Dalam is an environment with low levels of uranium oxide (and fluorine) in the percolating water, so it is very hard to see how Despott’s molar could have accumulated so much of it within just 7000 years. As Mifsud sums up:
The dating to the Neolithic in the 1964 Report could not be sustained in the face of the 13 ppm reading … Despott’s molar has survived to tell its tale … Its fluorine and uranium content ranks it contemporaneous with the fossil fauna of the Cervus Layer.75
As one might expect, knowing all the facts, this uranium oxide result was never published. The problem posed by the chemical results to the Neolithic date of first human settlement favoured by orthodox theory and confirmed in the 1964 Report has been efficiently dealt with by archaeology by simply ignoring the disturbing fluorine and uranium results whilst focusing only on a highly selective group of results from unreliable nitrogen assays. In consequence, until Mifsud rooted them out from the pages of the Natural History Museum’s Green Book and published them in Dossier Malta in 1997, neither the fluorine nor the uranium levels of any of the Ghar Dalam teeth were known outside the narrow circle of the two museums.
‘Adjustments’
In Dossier Malta, Mifsud claimed that the inconsistency between the uranium and fluorine results on the one hand and the high nitrogen result on the other could be explained by forgery. He took photographs of the Green Book during his visit to the Natural History Museum and spotted that there appeared to be two layers of ink in the box containing the nitrogen result for Despott’s molar. The bottom layer gave’. 8 per cent’ (i.e. 0.8 per cent, but without the zero). The top layer, in a different shade of ink, added a 1 and 5 to this result to give ‘1.85 per cent’.