New Evidence on the Antiquity of Piltdown Man

Dr. Kenneth P. Oakley, Department of Geology, British Museum (Natural History)

Dr. C. Randall Hoskins, Department of the Government Chemist

Nature March, 1950 xx

[379] Few, if any, fossils have given rise to more controversy than the remains discovered by Charles Dawson in gravel at Piltdown, near Fletching, Sussex, and described in 1913 by Sir Arthur Smith Woodward as representing a new genus and species of man, Eoanthropus dawsoni.

The geological age of 'Piltdown Man' was a matter of dispute from the first; moreover, this problem has latterly become linked with the question as to whether the thick but essentially human cranial bones and the remarkably ape-like lower jaw (and canine tooth) might be a chance association and represent two creatures of different geological ages. The present investigation, using the fluorine method for determination of the relative antiquity of fossil bones, has shown that the cranial bones and jaw-bone are of the same age, at the earliest Middle, more probably early Upper, Pleistocene.

Earlier Evidence

The Piltdown gravel is a thin remanié deposit in a terrace about 50 ft. above the River Ouse, containing fossil mammals of two distinct ages: a derived Villafranchian series ("Upper Pliocene" of earlier authors, now classed as Lower Pleistocene), and a later Pleistocene group., in part at least contemporary with the gravel. The "contemporary" group was classed by Dawson and Woodward 1 as "early Pleistocene", but according to more recent authorities it is not earlier than Middle Pleistocene.

To which of these two groups does Eoanthropus belong? This might be thought a simple question to decide by comparison of the states of preservation of the various remains. Yet examination of the specimens from this very point of view has led different authorities to diametrically opposed conclusions. In fact, anatomists regarding the mandible and canine tooth as anthropoid have felt free to place these with the Villafranchian group, and to dismiss the cranium as later Pleistocene and referable to Homo sapiens (cf. Marston 2 ). This hypothesis is permissible in so far as mammalian remains of two ages (for example, "Stegodon", Villafranchian; and Castor fiber, post-Villafranchian) occurred with Eoanthropus at the base of the dark gravel (ref. 3, pp. 83-85), [3] but it would be erroneous to claim that none of the cranial fragments was closely associated with the mandible. Although the fragments of skull-cap were for the most part recovered loose on spoil heaps, Dawson states (ref. 1, p. 121) that Smith Woodward dug out a small portion of the occipital bone "from within a yard of the point where the jaw was discovered, and at precisely the same level".

In reading their first paper, Dawson and Woodward (ref. 1, pp. 123, 143) stated their view that the human skull and mandible, being practically unrolled, were contemporary with the gravel and therefore probably "Pleistocene". However, in the discussion which followed, Sir Arthur Keith 4 argued that the skull should be assigned to the "Pliocene" group. He was influenced, he said, by the fact that in the Heidelberg jaw, of early Pleistocene date, the region of the chin was essentially human, whereas the Piltdown mandible showed simian characters. E.T. Newton 5 said that the mineralized condition of the skull bones pointed to their being of "Pliocene" age. In replying to the discussion, Dawson admitted this possibility, for he said (ref. 1, p. 151): ". . . the occurrence of certain Pliocene specimens in a considerably rolled condition, suggested a difference as to age, but not to the extent of excluding the possibility of their being coeval. The rolled specimens may have entered the stream further up the river than the human remains . . . . Then again the skull might have been surrounded by some colloid material [that is, clay] which preserved it in its passage from some earlier deposit." But in the second paper (ref. 3, p. 80) Dawson wrote: "Putting aside the human remains and those of the beaver, the remains all point to a characteristic land fauna of the Pliocene age; and though all are portions of hard teeth, they are rolled and broken. The human remains on the other hand, although of much softer material, are not rolled, and the remains of beaver are in a similar condition. It would therefore seem that the occurrence of these two individuals belongs to one of the periods of reconstruction of this gravel."

In 1935, Dr. A.T. Hopwood 6 reconsidered the evidence, and, largely on the basis of the state of preservation of the specimens, concluded that Eoanthropus belonged to the Villafranchian or Lower Pleistocene faunal assemblage. He pointed out that the absence of the indications of rolling was unreliable as a criterion, for one of the teeth of "Stegodon" (Elephas cf. planifrons ), undoubtedly a member of the derived group, was practically unrolled. The evidence then available, he said, justified the statement that "Piltdown Man is the oldest human fossil yet discovered" (ref. 6, p. 57).

In recent years many anthropologists have held the view that, so long as the date remained uncertain, this material was better placed in a suspense account, for the anatomical features of Eoanthropus (assuming the material to represent one creature) are wholly contrary to what discoveries in the Far East and in Africa have led us to expect in an early Pleistocene hominid. However, if one could at least be certain of the contemporaneity of the parts associated under the name Eoanthropus, and of their precise geological age, the number of possible interpretations would be much reduced.

The Fluorine-Dating Method

It has long been known that fossil bones accumulate fluorine in the course of time. The major constituent of bones and teeth is hydroxyapatite, which acts as a natural trap for wandering fluorine ions. Bones absorb fluorine from the ground-water, and it becomes fixed in their substance as fluorapatite by a process of ionic exchange. Owing to the porous texture of bones (and teeth), t his progressive alteration is not [380] confined to the surface, but usually proceeds more or less uniformly throughout the body of the material. It was suggested at the British Association meeting in 1947 7 that the percentage distribution of this element in the various bones and teeth from the Piltdown gravel might reflect their relative ages sufficiently clearly to throw some light on the major problem. Later, in response to a request from the Keeper of Geology at the British Museum, the Government Chemist agreed to undertake experimental work on the fluorine analysis of fossil bones.

After preliminary trials, Mr. R.H. Settle and one of us (C.R.H.), assisted by Mr. E.C.W. Maycock, adapted published methods of analysis to the exact determination of fluorine in very small samples of bone. A description of the analytical technique and a discussion on the limits of accuracy will be presented in a report on 'fluorine-dating' to be published later in the Bulletin of the British Museum (Nat. Hist ), and it is not necessary in this article to refer other than briefly to these matters.

The method was based on that originally devised by Williard and Winter 8 and afterwards modified by various authors. The fluorine in the bone was separated as hydrofluosilicic acid by distillation with perchloric acid in presence of a few beads of soft glass to serve as a source of silica. Suitable aliquots of concentrated distillate were titrated with dilute thorium nitrate solution, using Alizarin S as indicator. The solution of thorium nitrate was standardized against solutions of known fluorine content under identical conditions, particularly with regard to titrating to the same stage in the colour change of the indicator. In the majority of cases, the amount of fluorine in the distillate was sufficient for at least three aliquots to be titrated, and the average fluorine content, after making a small correction for a blank determination, was adopted. There was generally close agreement between the fluorine contents calculated from titrations of different aliquots from a given distillation.

Where possible, at least 20 mgm. of bone was used for fluorine determination; but in several cases it was necessary to rely on supplies of the order of 5 mgm. The errors of analysis naturally increase as the weight of sample decreases, but it is believed that with simple weights of 5 mgm., and upwards the error in the adopted values is not greater than ± 0.1 per cent of fluorine. For sample weights less than 5 mgm. the error may be ± 0.2 per cent of fluorine.

Approximate estimations of the iron and phosphate contents of the bones were made on residues of the samples remaining after the fluorine determinations. Colorimetric methods were used in each case (thioglycollic acid for iron, and ammonium molybdate followed by reduction of phosphomolybdate with stannous chloride for phosphate).

The fluorine-dating method was first applied to the Galley Hill skeleton 9. Briefly, it was shown that indigenous fossil bones in the Middle Pleistocene terrace gravels at Swanscombe contain around 2 per cent fluorine, those from Upper Pleistocene deposits in the same region around 1 per cent, and post-Pleistocene bones not more than 0.3 per cent; while the Galley Hill skeleton, although found in the Middle Pleistocene gravels, proved to contain only about 0.3 per cent fluorine, and was therefore clearly an intrusive burial, at earliest end-Pleistocene. The Swanscombe skull bones, on the other hand, discovered in these gravels by Mr. A. T. Marston in 1935-36, showed the expected 2 per cent fluorine.

Series of bones from other sites have been analysed, and the results show that the method, although limited in scope, is useful for differentiating fossil bones of diverse antiquity when they occur mixed together, provided that the specimens compared have similar matrices. It cannot be used to determine the relative antiquity of bones from widely separated localities. The method was ideally suited to the Galley Hill problem. There seemed reasonable hope that it would help to resolve the Piltdown enigma. Accordingly, in October 1948, Mr. W.N. Edwards, Keeper of Geology in the British Museum, authorized the sampling of Eoanthropus and associated mammalian bones and teeth. For the most part the samples were obtained by applying a dental drill to broken or worn edges of the specimens until a small but sufficient quantity of bone powder had been cored out. Where possible, powder from several drill holes in each specimen was mixed in order to ensure a representative sample. In view of the ferruginous nature of the deposits, it was thought advisable to determine the iron content of all the samples, but we found that there is no appreciable correlation of the Piltdown material between fluorine content and iron impregnation.

In the case of coasely porous bone, it is sometimes difficult to obtain a sample which is completely free from all contamination. The fluorine content of a contaminated portion of a bone will obviously be misleadingly low. It was therefore decided to determine the phosphate content of all samples, and to express the fluorine value of each sample as the percentage ratio of fluorine to phosphate (as P205). This procedure facilitates comparison of the fluorine contents of bones in which there has been variable contamination.

Application to the Piltdown Material

Every available bone and tooth from the Piltdown gravel and from neighbouring deposits was analysed, including seventeen samples of the Eoanthropus material. The results are shown in the accompanying table. For comparative purposes the mammalian remains have been grouped according to known or probable age. Their colour, degree of rolling or other physical states have been ignored in making this age classification, except in the case of some of the sub-fossil or recent specimens. The teeth of Mastodon arvernensis, Elephas cf. planifrons (= "Stegodon" auctt.) and 'Rhinoceros' cf. etruscus cannot be younger than Lower Pleistocene. All the rest of the Pleistocene material from the gravel is either certainly later (Cervus elaphus, Castor fiber ), or possibly later (Hippopotamus sp., Equus sp., Cervus sp.), than Villafranchian. Any such post-Villafranchian elements might theoretically be of first interglacial age (when according to Hopwood's terminology the "Middle Pleistocene" fauna began to appear); of great interglacial age (that is, Middle Pleistocene of all authors), or of last interglacial age (that is, early Upper Pleistocene of some authors, late Middle Pleistocene of others). Since distinctively Cromerian forms are absent and as there are indications that the deposit has been repeatedly re-worked, the post-Villafranchian material is probably partly Middle and partly Upper Pleistocene. A problem of classification typical of this site is presented by the so-called bone implement from Piltdown (P. 18) It is part of the femur of an elephant, judged on the basis of size to be that of a member of the Elephas meridionalis-antiquus group). It is conceivably Cromeran (first [381] interglacial), more probably later, but can safely be classed as "possibly Middle or Upper Pleistocene".

Analysis of Fossil Materials from Piltdown

 

Fluorine %

P205 %

% F_ x100

% P205

Iron

%

Eoanthropus I

       

P.1 L. parietal -frontal (E.590) †

0.1 (2) *

21

0.5

7

P.2 L.. temporal (E.591)

0.4

18

2.2

7

P.3 R. parietal (E.592)

0.3

17

1.8

6

P.4 Occipital (E.593)

0.2

28

0.7

5

P.5 R. mandibular ramus (E.594)

0.2 (5)

20

1.0

6

P.17 Canine (E.611)

< 0.1

27

0.4

Trace

P.42 Molar (rm 1) (E.594)

< 0.1

23

0.4

Trace

         

Eoanthropus II

       

P.30 R. frontal (E.646)

0.1

13

0.8

12

P.31 Occipital (E.647)

0.1 (2)

17

0.6

17

P.32 Molar (lm 1) (E.648)

0.4 (2)

30

1.3

Trace

         

Other Mammalian Remains

Lower Pleistocene

       

P.6 Molar, Mastodon cf. arvernensis (E.595)

1.9

23

8.3

5

P.7 Molar, Elephas cf. planifrons (E.596)

2.7 (2)

33

8.2

3

P.8 Molar, Elephas cf. planifrons (E.597)

2.5

34

7.4

1

P.23 Molar, Elephas cf. planifrons (E.620)

3.1

39

7.9

4

P.25 Molar, Mastodon arvernensis (E.622)

2.3

36

6.4

4

P.26 Premolar, 'Rhinoceros' cf. etruscus (E.623)

2.0

24

8.3

6

         

Possibly Middle and Upper Pleistocene

       

P.9 Molar, Hippopotamus sp. (E.598)

0.1 (3)

37

0.3

3

P.10 Premolar, Hippopotamus sp. (E.599)

l.1 (3)

29

3.8

5

P.11 Antler, Cervus elephus (E.600)

l.5 (3)

28

5.4

3

P.12 Metatarsal, Cervus sp. (E.601)

0.1

27

0.4

4

P.13 Molar, Equus sp. (E.602)

0.4 (3)

25

1.6

2

P.14 Molar, Castor fiber (E.603)

0.1

30

1.3

1

P.18 Femur, Elephas cf. antiquus (E.615)

1.3 (3)

30

4.3

2

P.19 Indet. bone from basal clay (E.616)

1.4

33

4.2

1

P.21 Incisor, Castor fiber (E.618)

0.1

27

0.4

10

P.22 Mandible, Castor fiber (E.619)

0.3

18

1.7

6

P.24 Frag. of enamel, Elephas sp. indent (E.621)

0.8 (3)

36

2.2

1

         

Holocene or Pleistocene

       

P.36 Tibia, Cervus sp. (E.1383)

< 0.1

35

0.3

1

P.37 Caprine molar (E.1384)

0.3 (3)

22

1.4

2

P.39 Bovine long-bone (E.1385)

0.1 (3)

30

0.3

9

P.40 Indet. bone (sub-fossil) (E.1386)

0.1

30

0.3

Trace

P.41 Indet. bone (sub-fossil) (E.1287)

0.3

42

0.7

2

         

Holocene (Recent)

       

P.33 Fragment of fresh bone from soil

< 0.1

33

0.3

Trace

P.34 Pelvis, Bos taurus

< 0.1

24

0.4

4

P.35 Metatarsal, Bos taurus (E.1388)

< 0.1

27

0.4

2

P.38 Ungual phalange, Bos taurus (E.1389)

0.3

32

0.9

5

* Where more than one determination of fluorine content has been made, indicated by the number in brackets, the value recorded is the average.

† The register numbers of specimens in the Department of Geology, British Museum (Nat. Hist.), are given in brackets after the description.

In attempting to interpret the analytical results, it is important to note that there is no significant difference in the rate of absorption of fluorine by bone and by dentine. The fluorine content of the mandibular ramus of Eoanthropus, for example, ranged from less than 0.1 to 0.3 per cent, while that of a molar embedded in this jaw-bone was less than 0.1 per cent. If there is differential absorption, the slight advantage is with bone. There are indications that enamel is more resistant to absorption of fluorine than dentine. The analysis of teeth, with two exceptions, were based on samples which were either wholly dentine, or which included a substantial proportion of dentine. The two exceptions were samples of enamel (P.6, P.24).

On the evidence of their state of preservation, the molar and premolar of Hippopotamus were placed by Hopwood (ref. 6, p. 49) with Eoanthropus in the Lower Pleistocene group. He noted, however, that Hippopotamus had never been recorded in a Red Crag association; so that on general grounds it would appear that these teeth more probably belong to the later group 10. The molar, the fluorine content of which is closely comparable with that of Eoanthropus, shows unique preservation (P.9). Whereas its enamel (0.1 per cent fluorine) is practically unaltered, its dentine is stained blackish-brown throughout and contains 0.05 per cent fluorine. An X-ray powder diffraction photograph showed that this blackened dentine consists of hydroxyapatite with slight admixture, possibly, of calcium sulphate. Analysis indicated 7 per cent iron. A probable interpretation of the specimen is that at some stage of fossilization the hydroxyapatite prisms in the dentine became coated by iron sulphate which inhibited fluorine absorption. The teeth of Eoanthropus are in striking contrast to this Hippopotamus tooth, and indeed to all the associated fossil animal teeth. Drill holes in the canine and in the molars of Eoanthropus revealed–most unexpectedly–that below an extremely thin ferruginous surface stain their dentine was pure white, apparently no more altered than the dentine of recent teeth from the soil.

Comparison of the fluorine values of the specimens attributed to Eoanthropus and of the bones and teeth the geological age of which are certain leaves little doubt that: (1) all the specimens of Eoanthropus, including the remains of the second skull found two miles away, are contemporaneous; (2) Eoanthropus is, at the earliest, Middle Pleistocene.

There can no longer be any question of Eoanthropus belonging to the Villafranchian group; whether it is Middle Pleistocene or later is arguable. That the figures scarcely provide any differentiation between Eoanthropus and recent bones requires some explanation; but at least it serves to emphasize the probably enormous time-gap separating the former from the Lower Pleistocene material.

The wide range of fluorine content in the post-Villafranchian material (0.1-1.5 per cent) is consistent with the suggestion that this age group is composite. Although none of these specimens is markedly water-worn, the gravels could have been reconstructed at several dates without the component materials travelling far. It is interesting to recall that Dawson (ref. 3, p. 86) considered that the remains of beaver (Castor ) were the only fossils from the Piltdown pit which could be counted as contemporary with Eoanthropus.. This is precisely the conclusion which one would draw from the fluorine results.

The eroded (and afterwards worked) fragment of elephant femur (P.18, with 1.3 per cent fluorine) was not found in situ, but appears to have come from clay below the gravel, where it would surely have been a derivative from an older deposit. The red-deer antler (P.11, with 1.5 per cent fluorine) was found some distance from the skull site, and may have been preserved in a patch of gravel also representing an earlier phase of the Pleistocene. The fluorine results are, in fact, so consistent with the known or probably relative ages of the mammalian fossils in the Piltdown mélange, that it now appears justifiable to regard Eoanthropus and Castor fiber as the latest elements in the mixture, and to ascribe [392] them to the period immediately preceding the final re-arrangement of the gravel, since which time free fluorine ions have apparently been remarkably deficient in the ground-water.

Geological Evidence: Conclusion

From the palaeontological data alone, it is not possible to decide whether the final settlement of the Piltdown gravel took place during Middle or early Upper Pleistocene times.

It was pointed out by Clement Reid 11 that these gravels rest on a low plateau surface (100-120 ft. above O.D.) which was extensively developed in the Weald, but which was nowhere covered by marine deposits of the period of the submergence represented by the Goodwood raised beach (135 ft. above O.D.). In any event, it seems unlikely that this surface existed before the base levelling associated with the close of the great interglacial period. Furthermore, if the Piltdown gravel is viewed as a river terrace deposit, its position about fifty feet above the River Ouse 12 places it in a group grading with the Main Monastrian sea-level. From the temperate character of even the latest faunal elements in the Piltdown faunal mélange, it is probable that Eoanthropus lived under interglacial conditions, although the final resorting of the gravels may have been brought about by periglacial soil-flow (solifluxion). The question of the precise geological age of the Piltdown gravel is open to further inquiry, but taking the balance of available evidence, Eoanthropus may be provisionally referred to the last warm interglacial period (Riss-Würm interglacial; that is, early Upper Pleistocene, although here it should be noted that some authorities count Russ-Würm as Middle Pleistocene).

We wish to record our thanks to the Government Chemist, Dr. G.M. Bennett, and to the Keeper of Geology, British Museum, Mr. W.N. Edwards, for their co-operation, and for permission to publish the relevant portions of this communication.

______________________________________________________________________________

1 Dawson, C., and Woodward, A. S., Quart. J. Geol. Soc. London, 69, 117 (1913).

2 Marston, A T., Geol. Assoc. London, Circular, No. 483, 1 (1948).

3 Dawson, C., and Woodward, A. S., Quart. J. Geol. Soc. London, 70, 82 (1914).

4 Keith, A., Quart. J. Geol. Soc. London, 69, 148 (1913).

5 Newton, E.T., Quart. J. Geol. Soc. London , 69, 151 (1913).

6 Hopwood, A.T., Proc. Geol. Assoc. London, 46, 46 (1935).

7 Oakley, K.P., Advancement of Science, 4, 336 (1948).

8 Willard, H.H., and Winter, O.B., Indust. Eng. Chem. (Anal. Edit.), 5, 7 (1933).

9 Oakley, K.P., and Montagu, M.F.A., Bull. Brit. Mus. (Nat. Hist.), Geol., 1, 2 (1949).

10 Curwen, E.C., "Archaeology of Sussex," 38 (London: Methuen, 1937).

11 Reid, C., Quart. J. Geol. Soc. London, 69, 149 (1913).

12 Edmunds, F.H., Abs. Proc. Geol. Soc. London, No. 1457, 39 (1950).