Middleton, Carnot and the initial formulation of a fluorine dating method

The Italian chemist Domenico Morichini appears to have been the first to discover that fluorine was present in fossil bone. He analysed the chemical composition of a fossilized elephant tooth found by the Italian natural philosopher Carlo Lodovico MorozzoFootnote ¹⁸ and published a paper in 1805 describing his results.Footnote ¹⁹ The German chemist Martin Heinrich Klaproth was one of the first scientists to examine Morichini's discovery,Footnote ²⁰ and soon thereafter the Swedish chemist Jöns Jakob Berzelius confirmed the presence of fluorine in living and fossil bone.Footnote ²¹ Edmond Frémy, professor of chemistry at the Muséum national d'histoire naturelle in Paris, conducted a more extensive series of studies on the chemical composition of fossil bone in the 1850s. He confirmed the presence of fluorine, but the quantities were small and extremely difficult to measure.Footnote ²² The idea that the fluorine content of fossil bones could be used as a means of dating them was first suggested by James Middleton, professor at University College London and fellow of the Geological Society of London. Middleton was drawn to this question by reading Berzelius's studies of the fluorine content of fossil and recent bone. Early in 1844 he began a series of experiments on the subject. Middleton started from the observation that fossil bone appeared to contain a greater quantity of ‘fluoride of calcium’ than recent bone. Chemists disagreed, however, about the means by which fluorine was absorbed by bone. Middleton came to suspect that fluorine was present in groundwater. If that were the case then ‘its accumulation in fossil bones would be found to be the result of infiltration’.Footnote ²³ In order to test his hypothesis, Middleton looked for fluorine in a water pipe from a coal mine, a stalactite, the deposits left in a tea kettle used to boil water, and several other objects. In all cases fluorine was found, in differing quantities, but this sufficed to convince Middleton that the source of the fluorine in fossil and living bone was groundwater.Footnote ²⁴

Middleton realized that if fluorine were absorbed by bone from groundwater then the amount of fluorine in a bone should increase over time. This idea prompted him to undertake a series of experiments designed to ascertain ‘the law by which fluoride of calcium becomes augmented or developed in fossil bones, as, should this be established, an important step would, I conceived, be thereby made towards the ascertainment of geological time’. The fluorine content of fossil bone might be used as a means of determining the relative, if not the absolute, age of fossils. This supposition led Middleton to investigate the amount of fluorine present in certain aqueous geologic deposits of different ages. He found that fluorine was present in all cases ‘from the most recent deposit down to the old red sandstone, and that it is present in the older in larger proportion than in the newer beds’.Footnote ²⁵ From this discovery Middleton concluded that he might be able to show ‘that the relative geological age of rocks may be estimated by the proportion of fluoride of calcium which they contain’.Footnote ²⁶

Middleton then used fossil specimens from different geological ages to see if their fluorine content increased with their geological age. He obtained horse, camel and alligator fossils from the geologist Hugh Falconer, who had excavated the fossils from the Siwalik Hills in India. He also examined the bones of an Iguanodon, a Miocene sea urchin, a two-thousand-year-old human skull found in Greece (the age of the skull was known from a coin placed in its mouth) and the skull of an Egyptian mummy. In a paper published in the Philosophical Magazine Middleton gave the quantities of fluoride of calcium, as well as other chemicals, found in each specimen, but drew few conclusions from these results, although the quantity of fluorine in each specimen was generally proportionally greater the older the specimen.Footnote ²⁷ However, in a paper read before the Geological Society of London, Middleton gave the results for recent bone, the ancient Greek skull, the Siwalik fossils and a fossil Anoplotherium bone (a mammal from the Eocene period), showing how the amount of fluorine in each specimen increased according to its geological age. Moreover, as the age of the Greek skull was known to be two thousand years, Middleton boldly used the relative ratios of fluorine in each specimen to estimate their absolute ages. Thus he suggested that the Siwalik fossils should be 7,700 years old and the Anoplotherium 24,200 years old.Footnote ²⁸

Middleton's research was the first to suggest a means by which bone absorbed fluorine and from this to establish the basic principles of a method to date fossil bone. Moreover, Middleton recognized the significance of his work in relation to one of the pre-eminent problems of geology and palaeontology in his day. He proclaimed before the Fellows of the Geological Society of London that the ‘accumulation of fluoride of calcium in fossil bones constitutes a very interesting and important subject of inquiry in reference to Geology, since it seems to involve the element of time, so interesting in all geological investigations’. Knowing the significance of this problem, Middleton began his experiments ‘to ascertain the proportion of fluoride of calcium in bones that had been preserved for various periods, with a view to infer, if possible, from the mineral condition, the relative ages of the specimens’.Footnote ²⁹ Yet Middleton did not expand upon his initial researches and published no further papers on the subject. The basic concepts that Middleton outlined for the use of fluorine to date fossil bone were not developed further until half a century later.

Middleton's research attracted little attention and it appears that no further efforts were devoted to using the fluorine content of fossil bones to determine their ages until the 1890s, when the French archaeologist Emile Rivière resurrected the technique. During the 1870s and 1880s Rivière excavated a number of Neolithic and Palaeolithic sites in France that in some cases produced stone artefacts in association with the fossil remains of mammoths and other Pleistocene fauna. This was an exciting time in French prehistoric archaeology: Louis Lartet had discovered the fossil remains of Cro-Magnon Man at Les Eysies in 1868 and Neanderthal fossils were discovered at Spy in Belgium in 1886. Rivière's excavations were part of a widespread effort by European archaeologists to uncover prehistoric artefacts, and especially human fossil remains, from the Ice Age or earlier. It was generally accepted that the approximate age of human artefacts or skeletal remains could usually be judged from the animal fossils with which they were found associated when excavated. But it was also recognized that such associations could be deceptive, since there were natural mechanisms, such as floods, that could mix objects from different stratigraphic layers and different geological ages. Moreover, the custom of burying the dead also led to human remains being accidentally mixed with the bones of much older animal fossils.

Rivière encountered just such an instance in the course of excavating at Billancourt, on the outskirts of Paris, where he discovered the bones of Quaternary animals along with Palaeolithic artefacts and portions of two human skeletons. However, Rivière noted that the fossils' appearance, as well as their density and general physical characteristics, differed greatly from the animal bones found with them. These facts convinced Rivière that the human remains were undoubtedly much younger in age.Footnote ³⁰ Several prominent archaeologists and palaeontologists disagreed with Rivière's conclusion, most notably Jean Louis Armand de Quatrefages, professor of anthropology at the Muséum national d'histoire naturelle. There was no clear way to settle the debate and the dispute continued for the next ten years. Rivière faced the opposite problem in 1892, when he unearthed two human skeletons from a cave at Baoussé-Roussé near Menton. He had been excavating there since 1870 and found numerous animal fossils dating from the late Quaternary period along with Solutrean, Mousterian and Magdalenian flint tools.Footnote ³¹ Because the deposits in the cave showed no evidence of being disturbed, Rivière was adamant that the animal fossils and the human artefacts and skeletons were deposited in the cave at the same time and were therefore of very great age.Footnote ³² He suggested that the human skeletons belonged to the race of Cro-Magnon,Footnote ³³ but once again there were sceptics, notably the anthropologist René Verneau, who rejected this conclusion and suggested that the human remains dated from the Neolithic and were therefore not so ancient.Footnote ³⁴ Rivière vigorously defended the late Quaternary age of the human remains based upon the geological evidence, but had already taken steps to discover a way of determining with certainty the ages of the Billancourt and Baoussé-Roussé skeletons.Footnote ³⁵

In October 1885 Rivière deposited a pli cacheté with the Académie des sciences regarding his opinion about the human skeletons found at Billancourt, with instructions that it should be opened at the completion of some new investigations he was about to start on the chemical composition of fossil bones of different ages from various geological deposits.Footnote ³⁶ He got this idea from reading works published in 1860 on this subject by Alphonse Milne-Edwards and Achille-Joseph Delesse. In order to carry out the chemical analysis of the Billancourt bones, Rivière enlisted the assistance of Adolphe Carnot, professor of chemistry at the Ecole des mines. Carnot was an ideal choice in many ways, since he was a chemist and mining engineer with experience in geology. Carnot immediately focused on the fluorine content of the fossils as the subject of his investigations, in part from reading Edmond Frémy's 1855 work on the subject.Footnote ³⁷

Carnot, who gives no indication of having read Middleton, undertook a much more extensive study of fossils of different geological age than had the London professor, obtaining many of his fossil specimens through the assistance of Albert Gaudry, professor of palaeontology at the Muséum.Footnote ³⁸ The procedures of chemical analysis and knowledge of palaeontology and geological strata had all expanded considerably since Middleton's time, and Carnot made good use of this new knowledge. He began by chemically analysing more than fifty fossil bones (the oldest fossils belonging to the Silurian period) and twenty modern bones, then compared the proportion of ‘phosphoric acid’ (PhO₅) to fluorine in each fossil. Carnot described in some detail the procedures he followed to isolate and measure the quantity of fluorine and other substances of interest in each specimen, clearly a laborious and demanding process.Footnote ³⁹

One objective of these investigations was to compare the chemical composition of modern and fossil bone in order to determine the chemical processes that caused fossilization – processes not then well understood.Footnote ⁴⁰ This was important for the interpretation of the cause and significance of the presence of fluorine and other chemicals in fossil bones. Carnot gave charts listing the amounts of various chemicals found in fossils from each geological period, with a clear implication that while the chemical composition of fossils could vary greatly, the amount of fluorine generally increased with the age of the fossil.Footnote ⁴¹ Carnot explained the presence of fluorine and some other chemicals in fossils by suggesting these chemicals existed in groundwater and gradually infiltrated the bone buried in these sediments. To test this hypothesis, Carnot conducted several experiments where he immersed modern bone in a fluorine solution and tested the amount of fluorine absorbed by the bone at regular intervals, which showed that fluorine was indeed absorbed by bone.Footnote ⁴² From this he drew the significant conclusion that similar continuous action over a very long period could result in much greater concentrations of fluorine in bone, exactly what his analysis of fossilized bones indicated.Footnote ⁴³

There was, however, one problem with the use of fluorine to determine the age of a fossil. Carnot noted that the amount of fluorine in fossils of the same geological age but taken from different geographical localities showed ‘very great variation’.Footnote ⁴⁴ He attributed this to differences in the chemistry of the local groundwater and soil.Footnote ⁴⁵ Yet this had profound consequences for the usefulness of any fluorine dating method, since it meant that no general method could be formulated to assign a geological age to a fossil solely on the basis of its fluorine content because that quantity would vary from place to place.Footnote ⁴⁶ Carnot's investigations did nevertheless have immediate practical consequences for prehistorians, because the significant difference in fluorine content between modern bone and Pleistocene bone meant that it should be possible to identify modern bones that had become accidentally mixed with older animal fossils, as long as both sets of fossils came from the same location.Footnote ⁴⁷ Carnot was now ready to address the issue of the human and animal fossils encountered by Emile Rivière at Billancourt. Carnot analysed two fragments of animal bone from Billancourt, as well as a human tibia sent to him by Rivière. The results were dramatic and convincing. The animal fossils contained much more fluorine than the human bone. The amount in the animal bones corresponded very well with other Quaternary fossils while the human bone was consistent with modern bone. From this Carnot concluded that the human remains were much younger than the animal fossils, thus their association in the geological deposits at Billancourt was accidental, just as Rivière had asserted all along.Footnote ⁴⁸

Carnot argued that ‘this new method of control may often be useful in the examination of problems relative to the antiquity of man’ since human and animal bones buried in the same deposits at the same time should always contain the same amounts of fluorine, but if fossils of different ages had become mixed in alluvial deposits or in caves then this would be shown by the differences in their fluorine content.Footnote ⁴⁹ Both Carnot and Rivière published accounts of these investigations, but while Carnot gave details of his chemical analysis and its results, he did not discuss the geological or archaeological details of the Billancourt fossils, although he did discuss the potential application of his method to prehistoric research. Rivière only mentioned the most basic results of the chemical analysis conducted by Carnot and focused instead on showing how these results confirmed his earlier conclusions about the Billancourt fossils based on his palaeontological and archaeological interpretation of the human and animal remains.Footnote ⁵⁰

Both Carnot and Rivière saw the potential for this method to resolve a common problem in prehistoric archaeology. However, despite publication of the results of their investigations in prominent scientific journals, only a few other archaeologists utilized Carnot's method. In Mexico the biologist Alfonso Luis Herrera had a human mandible discovered on the shore of Lake Chalco and a fossil horse skull from the same deposit tested for their fluorine content. The tests showed that the ‘Xico mandible’ and the extinct horse were contemporaneous.Footnote ⁵¹ Two years later, Thomas Wilson, curator of the Department of Prehistoric Anthropology at the Smithsonian Institution, announced the results of his analysis of the controversial ‘Natchez Man’ specimen discovered in 1845 by Montroville Wilson Dickeson, a physician and amateur archaeologist. Dickeson discovered the human pelvis near Natchez, Mississippi in geological deposits that also contained numerous extinct Pleistocene animal fossils. Dickeson presented this remarkable collection to the Academy of Natural Sciences of Philadelphia in 1846, noting that the human bone was found directly below the animal fossils and was also clearly fossilized.Footnote ⁵² Considerable debate surrounded the question of whether this was evidence that humans were present in America along with the mastodon and other Ice Age mammals. Wilson saw Carnot's fluorine analysis as a possible way to settle the question. Wilson obtained samples of the Natchez fossils and gave them for analysis to Robert Lawrence Packard, a mineralogist at the Smithsonian. Although the amount of fluorine was somewhat less in the human bone, Wilson believed they were about the same age, dating from a period midway between the present and the Quaternary period.Footnote ⁵³ Wilson and Packard also published an English translation of Carnot's 1893 paper, thus making his means of analysing fossil bone accessible to American scientists.Footnote ⁵⁴

The Dutch chemist Jakob Maarten van Bemmelen also used Carnot's method of fluorine analysis to test the famous Pithecanthropus erectus fossils discovered on Java by Eugène Dubois between 1891 and 1892. From the anatomy of the fossils and the apparent late Pliocene or early Pleistocene geology of the site, Dubois was convinced that Pithecanthropus represented the intermediate link between apes and humans. Many of Dubois's colleagues were sceptical of this interpretation of the fossils. One problem was Dubois's assertions about their geologic age.Footnote ⁵⁵ As a result, Dubois despatched a fragment of Stegodon bone found in the same deposits as the Pithecanthropus fossils, to test its fluorine content so that it could be compared with those of European fossils given by Carnot. The amount of fluorine present in the Stegodon bone was consistent with late Pliocene fossils from Europe.Footnote ⁵⁶ Yet this analysis had little impact on the debate over the Pithecanthropus fossils. Indeed, despite the promise and the apparent success of Carnot's fluorine dating method in the cases of the Billancourt, Xico and Natchez human fossils, the method never became widely used and fell into relative oblivion during the early decades of the twentieth century.

Kenneth Oakley and the modern fluorine dating method

The fluorine dating method had all but disappeared as a tool in palaeoanthropological research during the early decades of the twentieth century. The British geologist and palaeoanthropologist Kenneth Oakley almost single-handedly revived the method in the 1940s and transformed it into a useful and widely used technique. Oakley studied geology and anthropology at University College London and in 1935 became assistant keeper of palaeontology at the Natural History Museum. During the Second World War Oakley was attached to the Geological Survey of Great Britain, where he conducted research on the use of phosphates in fertilizer as well as the impact of fluorine on dental health. This latter project brought Oakley into contact with Adolphe Carnot's research.Footnote ⁵⁷ The first occasion Oakley had to apply Carnot's method of analysing fossil bone arose from the First Pan-African Congress on Prehistory, held in Nairobi in 1947. Oakley had discussed the fluorine dating method; Louis Leakey, organizer of the congress, realized that Oakley might have the solution to a long-standing problem. Leakey had discovered two partial human crania at Kanjera near Lake Victoria and a human mandible at another site, Kanam, in 1934. Both were remarkably modern in their anatomy but came from very old geological deposits. Leakey took these fossils to be evidence that anatomically modern humans had evolved very early in Africa, but many of his critics viewed the fossils with suspicion.Footnote ⁵⁸ Since animal fossils had been found in the deposits containing the human skull and mandible, the fluorine test might determine if they were of the same age or if the human fossils were recent intrusive burials, as some believed.

Leakey provided Oakley with a sample from a Kanjera skull, along with animal specimens from the same deposit, as well as a late prehistoric human bone from the same region. Oakley submitted these samples to Henry James Walls of the Home Office Forensic Science Laboratory in Bristol for chemical analysis. Unfortunately, Walls found that the Kanjera fossils were all saturated with fluorine, the result of the extremely high concentrations of the element found in the Kanjera deposits due to their volcanic origin.Footnote ⁵⁹ While this first attempt to use the fluorine dating method was unsuccessful, Oakley was convinced of its potential value to palaeoanthropology. In fact, he already had two candidates for fluorine analysis in mind, the Galley Hill skeleton and the troublesome Piltdown Man fossils.Footnote ⁶⁰

The Galley Hill skeleton, already mentioned above in the context of the presapiens hypothesis, was discovered in 1895 in Kent in Pleistocene geological deposits. The skull of Galley Hill Man had been found seven years earlier. The skeleton was interesting because of its apparent great age and the fact that it was anatomically modern. The palaeontologist Edwin Tulley Newton, member of the Geological Society of London, published an account of the skeleton in 1895 in which he argued for the antiquity of the Galley Hill skeleton on the basis of the geological strata in which it was found and the stone implements excavated in the same strata. Newton also emphasized certain morphological features of the skull that indicated it might be very old.Footnote ⁶¹ There was, however, suspicion concerning the supposed age of the skeleton from the very beginning. John Evans, a respected and influential prehistoric archaeologist who had participated directly in the geological debate over the coexistence of humans and Ice Age mammals, doubted it was as old as Newton claimed. This view was shared by others who suspected it might be a relatively recent skeleton buried in a much older deposit.Footnote ⁶² The debate over the age of the Galley Hill skeleton grew during the early decades of the twentieth century. Prominent figures such as Arthur Keith supported its antiquity while other eminent scientists such as Wynfrid Laurence Henry Duckworth concluded it was a more recent intrusive burial.Footnote ⁶³ Oakley saw the fluorine dating method as a means of settling the debate. In 1948 he and Wilfred Norman Edwards, keeper of geology and palaeontology at the Natural History Museum, approached the Department of the Government Chemist in London seeking assistance in conducting a chemical analysis of the Galley Hill skeleton and the Pleistocene animal fossils from the same deposit. Three chemists – R. H. Settle, C. Randall Hoskins and E. C. W. Maycock – agreed to join the investigation.Footnote ⁶⁴

In order to ensure the reliability of the results of their analysis of the fluorine content of the Galley Hill collection, they first needed to

A series of twenty-two bone samples from the Galley Hill site were then chosen for analysis, including fossils from the Lower Pleistocene deposits where the human skeleton was found, as well as more recent fossils from Upper Pleistocene deposits taken from nearby gravel pits and samples from a Saxon skeleton from historic times.Footnote ⁶⁶ It was then necessary that the chemists obtain a ‘representative sample’ of bone for analysis from each specimen. In order to achieve this, the chemists drilled holes about five millimetres deep from three or four widely separated spots on each bone and the powder produced from each drill hole was then mixed together.Footnote ⁶⁷

Once samples had been obtained in this way from each specimen of bone they were subjected to analysis. The chemical procedures used to isolate and measure fluorine had advanced significantly from the techniques used by Carnot and Middleton. When Henry Walls had examined the Kanjera fossils he had used a new method of determining fluorine content devised by R. F. Milton, H. F. Liddell and J. E. Chivers.Footnote ⁶⁸ To determine the fluorine content of the Galley Hill fossils, Settle, Hoskins and Maycock used a slightly different method developed by Hobart Willard and O. B. Winter in 1933.Footnote ⁶⁹ Both methods were extremely time-consuming and exacting, severely limiting the number of specimens that could be reasonably and affordably analysed. This point was emphasized by Oakley, who remarked that ‘in order to be used accurately this method requires considerable technical experience’.Footnote ⁷⁰ Unlike the Kanam and Kanjera analysis, that of the Galley Hill fossils produced very good results. The fluorine content of the specimens showed conclusively that the Galley Hill human skeleton did not date from the Pleistocene, but was instead a recent burial.Footnote ⁷¹

The validity of the fluorine method was strengthened when another fossil, a human skull found in 1935 at Swanscombe in Kent, was also tested. The Swanscombe skull was found in association with Acheulean flint tools, thus placing it in the Lower Middle Pleistocene. The fluorine analysis of the skull showed that it was consistent with this early period.Footnote ⁷² The success of the fluorine dating method in resolving the Galley Hill debate and in confirming the antiquity of the Swanscombe skull led Oakley to think that it might finally resolve one of the most debated problems in human palaeontology, the age and validity of the Piltdown Man specimen. Between 1908 and 1912 Charles Dawson, a solicitor and amateur archaeologist, had discovered fragments of a modern-looking human skull along with a portion of a very ape-like mandible in a gravel pit in Piltdown Common in Sussex. The geologist Arthur Smith Woodward and anatomist Arthur Keith believed that the skull and jaw belonged to a species of early human they called Eoanthropus dawsoni. Some anthropologists believed that the skull and jaw fragments belonged to two separate creatures, whilst other scientists were sceptical of the supposed Lower Pleistocene age of the fossils.Footnote ⁷³ A further complication was introduced in 1946 when Alvan T. Marston, a London dentist with extensive experience in Palaeolithic archaeology, suggested that the Piltdown skull and jaw might not be of the same age. Oakley had discussed the use of the fluorine dating method for the Piltdown fossils in 1947 at the Dundee meeting of the British Association for the Advancement of Science: the success of the method with the Galley Hill and Swanscombe fossils prompted Oakley and his collaborators at the Department of the Government Chemist to submit the Piltdown fossils to fluorine analysis.Footnote ⁷⁴

In October 1948 Wilfred Norman Edwards authorized analysis of all the Piltdown fossils as well as animal fossils from the Piltdown gravels and nearby deposits. The oldest animal fossils were from the Lower Pleistocene and Villafranchian, while the remaining fossils were from the Middle to Late Pleistocene. Samples were taken from the broken or worn edges of each specimen in order not to damage the fossils. A dental drill was used extract about twenty milligrams of powder from each specimen.Footnote ⁷⁵ In this case Randall Hoskins conducted the chemical analysis. The Lower Pleistocene animal fossils contained two to three per cent fluorine and the remaining animal fossils contained less than 1.6 per cent fluorine. The Piltdown Man fossils, however, all contained only about 0.2 per cent of fluorine.Footnote ⁷⁶ From these results Oakley concluded first that the skull bones and the mandible were the same age, which resolved Marston's doubts. It was also clear that Eoanthropus dawsoni could not date from the Lower Pleistocene and provisionally Oakley dated it to the last interglacial period (the Riss-Würm) or Upper Pleistocene.Footnote ⁷⁷ Oakley admitted, however, that the amount of fluorine in the Piltdown Man specimens was so low as to be comparable with modern bone, a fact that ‘requires some explanation’, although it could only serve ‘to emphasize the probably enormous time-gap’ separating the Piltdown Man remains from the Lower Pleistocene.Footnote ⁷⁸ This small matter would become of considerable importance just a few years later. In the interim, Oakley felt confident in asserting that the fluorine results had not only settled the question of the age of Eoanthropus dawsoni but had also ‘considerably increased the probability that the mandible and the cranium represent a single creature’.Footnote ⁷⁹

The fluorine dating method had clearly shown its value to palaeoanthropology. Oakley now turned his attention to other human fossils that were of doubtful antiquity. He and Hoskins examined human skull fragments found in 1882 at Westley, near Bury St Edmunds. Although the fragments were found associated with Acheulean hand axes in Middle Pleistocene deposits, fluorine analysis proved that they were from a recent intrusive burial.Footnote ⁸⁰ The next specimens to be analysed were the human skulls excavated from the French Palaeolithic site of Fontéchevade in 1947 by Germaine Henri-Martin. The Fontéchevade skulls were at the centre of a debate over the validity of the presapiens hypothesis because, like the Swanscombe and Piltdown fossils, they seemed to be similar in their structure to modern human skulls but were from very old geological deposits.Footnote ⁸¹ At the request of Henri-Martin, Oakley had the skulls tested. While the results did not definitively settle the question of their age, they were consistent with the supposed Tayacian age of the deposits where the skulls had been found.Footnote ⁸² From this examination, Oakley also learnt that the fluorine dating method ‘only yields conclusive results when the age-difference between the groups under comparison is sufficiently great. Thus, it is unsuitable in all but the most favourable circumstances for the differentiation of, say, Mousterian from Upper Palaeolithic, Upper Palaeolithic from Mesolithic.’Footnote ⁸³ This did not undermine the validity of the method, but it did imply that it had its limits and that this should be realized by palaeoanthropologists.

Fluorine analysis also showed that fragments of a modern human skull found in association with a horn from the extinct aurochs (Bos primigenius) in Suffolk were modern and that a human skull found along the banks of the Thames in deposits containing stone tools of the Levalloisian type was not contemporaneous with those tools. The notorious Moulin Quignon jaw, found in Abbeville in 1863 in deposits containing numerous Palaeolithic implements, also succumbed to the fluorine dating method.Footnote ⁸⁴ Oakley received permission to analyse the famous fossil from the Musée de l'homme in Paris and the results showed with certainty that the jaw was not from the Palaeolithic.Footnote ⁸⁵ Within just a few short years Oakley and his colleagues at the Department of the Government Chemist had resolved doubts about more than half a dozen fossils, indicating that several were not at all prehistoric while strengthening the ancient status of several others. Moreover, they had demonstrated the power of the fluorine dating method to solve problems that had resisted solution in any other way.

Oakley's investigations using the fluorine dating method attracted the attention of Paul Fejos, director of research of the Wenner-Gren Foundation (originally the Viking Fund), while Oakley was lecturing on this research in New York City in 1950. Fejos and the Wenner-Gren Foundation provided financial support for many types of anthropological and archaeological research and Fejos offered Oakley funds to allow for the expanded use of the fluorine dating method. Oakley discussed this offer with Wilfrid Norman Edwards of the Natural History Museum and the Department of the Government Chemist, which quickly led to an agreement on a long-term collaborative project between the two institutions to apply the fluorine dating method to much larger numbers of fossils from around the world. Oakley and his colleagues received their first grant from the Wenner-Gren Foundation in 1952 to purchase equipment and allow for travel abroad.Footnote ⁸⁶ It is significant that Fejos not only supported Oakley's fluorine dating investigations, but was also among the first to recognize the potential value of American physical chemist Willard Frank Libby's radiocarbon (carbon-14) dating method and he encouraged the foundation to fund that research as well.Footnote ⁸⁷ These two dating methods quickly began to transform the study of prehistory.

The fluorine dating method, still relatively little known among anthropologists, suddenly made its dramatic mark in 1953 as a result of a new series of investigations of the Piltdown Man fossils. Oakley had analysed the fossils in 1949 but the research had not resolved all the questions surrounding the Piltdown specimen. Renewed doubts and suspicions about the fossils finally prompted Joseph Sidney Weiner, professor of physical anthropology at Oxford University, to inspect the fossils anew during the summer of 1953. Weiner enlisted the assistance of Oakley and the Oxford anatomist Wilfrid Le Gros Clark; they were soon obtaining some very interesting results. Oakley once again had the skull fragments and the mandible subjected to fluorine analysis. On this occasion Charles Frederick Fryd of the Department of the Government Chemist used a new and much more sensitive technique. A startling difference in the fluorine content now appeared, not achievable with the original technique used in 1949. Oakley and Fryd found that the Piltdown skull fragments contained about 0.1 per cent of fluorine while the mandible only contained 0.03 per cent.Footnote ⁸⁸ This meant that the skull could still be prehistoric, but the mandible was certainly modern. Other chemical analyses and a close inspection of the fossils provided further evidence not only that the skull and mandible were not the same age and were not very old, but also that they were not what they appeared to be. Weiner, Oakley and Le Gros Clark found convincing evidence that the jawbone had been tampered with and that the Piltdown fossils were an elaborate forgery.Footnote ⁸⁹

The disclosure that the Piltdown fossils were a fraud created an uproar and attracted huge attention, because of the prominence of the Piltdown fossils and half-century of often acrimonious debate that surrounded them. There was also a great deal of interest in the means employed to uncover the fraud. Many of the chemical methods used were relatively new, at least in their application to anthropological problems, but they clearly showed the power of chemical analysis and especially of the fluorine dating method for resolving problems in palaeoanthropology. Because of the Piltdown affair, many palaeoanthropologists learnt of the fluorine dating method and its possible uses. With substantial funding from the Wenner-Gren Foundation to expand their research and the pivotal role that the fluorine dating method played in solving the Piltdown problem, the way was prepared for its widespread application.

After learning of Oakley's application of the technique to the Galley Hill skeleton, the University of California–Berkeley archaeology professor Robert Fleming Heizer and Sherburne Friend Cook, professor of physiology, had already begun to apply the fluorine dating method to palaeoindian sites in California.Footnote ⁹⁰ Throughout the 1950s they used the fluorine dating method, along with other chemical methods they were developing to date fossil bone, to resolve doubts about human fossils that were of uncertain age.Footnote ⁹¹ From 1953 into the 1960s Oakley oversaw the fluorine dating of numerous specimens including the Broken Hill, Florisbad and Saldanha human fossils from Africa, as well as a number of European fossils.Footnote ⁹² Other anthropologists also began to adopt the method. The American anthropologist Carleton Stevens Coon used the fluorine dating method in 1951 to test bones collected during excavations at Hotu Cave in Iran, but also submitted samples of charcoal from the site to be tested using the new radiocarbon (carbon-14) dating method.Footnote ⁹³

This juxtaposition of dating methods is significant, because the fluorine dating method flourished only for a brief period, while the carbon-14 dating method continued to grow in importance. Indeed, the fluorine dating method had all but disappeared by 1970, although it continued to be used in specific situations where other methods could not be applied.Footnote ⁹⁴ A noteworthy indication of the fading importance of the fluorine dating method by 1970 is the fact that while a chapter devoted to the technique appeared in the 1963 edition of Don R. Brothwell and Eric S. Higgs's Science in Archaeology, it was not even mentioned in the 1970 edition. The major reason for the decline of the fluorine dating method was the fact that it did not provide absolute dates for the specimens analysed, but could only show whether or not a group of fossils were contemporaneous. Developed at the same time as the refinement by Oakley and his colleagues of the fluorine dating method, the carbon-14 technique could provide absolute dates for organic specimens and excavation sites (if organic material was present), if those specimens were less than about fifty thousand years old. However, by 1958 the potassium–argon (K–Ar) dating method had been developed. This approach provided absolute dates for prehistoric specimens and sites millions of years old, as long as these were located in volcanic deposits. The successful application of the carbon-14 and potassium–argon dating methods to prehistoric artefacts and sites meant that the fluorine dating method was no longer needed, except under very specific conditions.

Yet between 1949 and 1969 the fluorine dating method was used to analyse a substantial number of human fossil specimens from around the world and made a significant contribution to palaeoanthropology by finally settling doubts about the antiquity of a large number of specimens. As apparent in the case of Rivière, the ability to distinguish truly prehistoric human fossils from recent intrusive burials was becoming a major problem by the late nineteenth century, when substantial numbers of archaeologists excavated sites seeking human bones or artefacts dating from the Pleistocene or earlier. Intrusive burials and the accidental association of bones of different ages through the disturbance of geological deposits created doubts about many supposed Palaeolithic human remains. The fluorine dating method devised by Carnot was the only available means that definitively resolved such cases. The method was also important in the history of palaeoanthropological research because it was a significant advance in the search for an absolute dating method. For the first time, the stratigraphic method used to obtain relative dates for archaeological and geological sites was supplemented by an entirely new chemical means to date fossil specimens. Although the fluorine dating method could not provide the absolute dates desired by palaeoanthropologists, the search for chemical or physical means of dating bones or artefacts did not end and was eventually successful, first with the carbon-14 dating method and later with other methods.

Moreover, the collaboration that this kind of research required, between anthropologists or archaeologists and chemists or physicists, so essential to the successful development and application of archaeological dating methods and the correct interpretation of their results, was already apparent in the relationship between Rivière and Carnot. It is even more visible in the close cooperation that arose between Oakley and chemists such as Hoskins and Fryd at the Department of the Government Chemist. Oakley brought his knowledge of palaeontology and geology to the interpretation of human fossil remains. Hoskins and Fryd applied their expertise and knowledge as chemists to exacting and difficult chemical analyses of fossils bones to produce numbers that had to be interpreted chemically before the results could give an understanding of the relative ages of the fossils under investigation. Oakley had to trust the skills and results of the chemists with whom he worked, while the chemists sometimes had to revise and improve the methods they employed to obtain the results anthropologists desired from them. Chemistry also benefited from this relationship. Hoskins and Fryd were able to publish a paper on a refined method for measuring quantities of fluorine in fossil bone that they had developed for the second round of Piltdown analyses.Footnote ⁹⁵ This kind of collaboration between scientists of different disciplines in palaeoanthropological research would become important not only with the development of new dating methods but also as modern palaeoanthropological expeditions and research teams became increasingly multidisciplinary in the 1970s. The development of the fluorine dating method, now a largely forgotten episode in the history of palaeoanthropology, not only was influential because of the results of its application in such cases as the Galley Hill skeleton and the Piltdown fossils, but also marks the beginning of significant changes in the way palaeoanthropological research would be conducted in the last half of the twentieth century. It reflects some of the challenges and rewards introduced by this new type of research.