Nature October 2, 1913

G. Elliot Smith

[131] It had been my intention not to add anything further in print to my preliminary note (Quart. Journ. Geol. Soc., vol lxix., 1913, p. 145) on the cranial cast obtained by Dr. Smith Woodward from his reconstruction of the Piltdown skull until I was in a position to make a full and comprehensive statement as to the precise significance of the information afforded by the cranial fragments as to the kind of brain possessed by the earliest known human inhabitant of Britain. But, although my investigations are now sufficiently advanced to permit me to undertake the writing of my report, it will be some months before it can be published; and in the meantime it is

most undesirable that the present widespread misunderstandings should be allowed to breed further trouble and confusion for those who are interested in the elucidation of Mr. Charles Dawson's momentous discoveries.

Recent events have made it difficult for those who have relied wholly upon what has appeared in print to form any accurate conception of the meaning and importance of the Piltdown skull-fragments. It is quite certain that they afford the first evidence we have obtained of a hitherto unknown group of the Hominidæ, so fundamentally distinct from all the early fossil men found in Europe as to be worthy of generic distinction–a "dawn-man" of a very primitive and generalised type. Certain features are so clearly ape-like as definitely to confirm the generally admitted kinship to the African anthropoid apes, as well as to distinguish Eoanthropus sharply and clearly from all other human remains. In other respects, however, there is a closer resemblance to the features of modern man than is found in the specialised group of Neanderthaloid palæolithic men. This curious association of features is not paradoxical, as some people pretend. The small and archaic brain and thick skull are undoubtedly human in character, but the mandible, in spite of the human molars it bears, is more simian than human. So far from being an impossible combination of characters, this association of human brain and simian features is precisely what I anticipated in my address to the British Association at Dundee (Nature, September 26, 1912, p. 125), some months before I knew of the existence of the Piltdown skull, when I argued that in the evolution of man the development of the brain must have led the way. "The growth in intelligence and in the powers of discrimination no doubt led to a definite cultivation of the æsthetic sense, which, operating through sexual selection, brought about a gradual refinement of the features." Just as the young child still uses its teeth for purposes of attack, so in the dawn of human existence teeth suitable for offensive purposes were retained long after the brain had attained its distinctively human status and had made the hands even more serviceable instruments for attack."

That the ape-like conformation of the chin region signifies the inability to speak is surely a patent fallacy. Articulate speech must have come while the jaws were still simian in character; and the bony changes that produced a chin were the result mainly of that process of refinement to which I have already referred, to the reduction of the teeth, which was part of the same process, and quite in a minor degree, to that process of growth and specialisation of the genio-glossal muscles which resulted from their use in speech.

A great source of misunderstanding will be got rid of if these obvious facts and the considerations based upon them be admitted.

In conclusion, I may answer many questioners by affirming that I still hold to every word of my preliminary note published in the Quarterly Journal of the Geological Society, as well as of the statements made in my lectures delivered before the Royal Dublin Society and the Manchester Literary and Philosophical Society last winter, and also to the facts demonstrated in my exhibit at the Royal Society's soirée in May.

Nature October 16, 1913

Arthur Keith

[197] The discovery of the fragments of the Piltdown skull has given rise to a problem of a new kind. In all former discoveries of the remains of ancient man the part of the skull actually found was intact,, or, if broken, a sufficient number of pieces were recovered to render reconstruction an easy task. In the case of the Piltdown skull, although the greater part of the bony walls of the cranial cavity were found, a large area of the forehead and along the middle line of the roof of the skull are still missing. The problem that has to be solved is: How much is missing? The solution of the problem, as Dr. Smith Woodward realised when he commenced his work of restoration, lies in the hinder or occipital Wall of the skull. The fragment which Dr. Smith Woodward himself discovered gives a definite index to the width of the right half of the occipital bone, and also to the width of the hinder or occipital part of the head.

Fig. 1–Occipital aspect of the brain-cast of the Piltdown skull as reconstructed by Dr. Smith Woodward. The parts missing in the skull are represented by vertical shading.

It is clear, then, that the first step in the reconstruction of the Piltdown skull must be an accurate adjustment of the parts which enter into the formation of the occipital wall. If a mistake is made in this initial step, then the error may become proportionately greater as one proceeds towards the region of the forehead. In my opinion, Dr. Smith Woodward has made a grave mistake in his restoration of the occipital region, and therefore the brain cast which he obtained from his reconstruction–the basis of Prof. Elliot Smith's preliminary note to the Geological Society–-does not give an accurate representation of either the size or general form of the brain of Piltdown Man.

The nature of the problem and the manner of .its solution will be made clear by the three accompanying figures. Fig. 1 represents the occipital aspect of the [198] brain cast obtained in Dr. Smith Woodward's reconstruction; Fig. 2, the same view in a reconstruction of the skull made by the writer; Fig- 3, the same view of a brain cast from the skull of an Australian native, with. a cubic capacity of 1460 cubic centimetres–rather a large skull for an Australian native. All these brain casts have been arranged on the same horizontal plane and drawn to the same scale. To facilitate comparison, they have been placed within squares of the same size. Three vertical lines are represented–the middle and two lateral lines. The lateral lines are 50 mm. apart from the middle line.

Fig. 2.–The same aspect from the reconstruction of the skull by the writer.

Fig. 3.– The same aspect of the brain cast from the skull of an Australian native–for comparison with figs. 1 and 2.

The leading principle which guides the task of reconstruction is symmetry–the right and left halves of the mammalian head and skull are approximately alike. When the test of symmetry is applied to the occipital region of the brain cast from Dr. Smith Woodward's reconstruction of the skull, it is at once seen that there is a great degree of discrepancy between the right and left halves; the amount which has to be added to the right half to make it approximately equal to the left is shown by the stippled line in Fig. 1. The discrepancy between the two halves is even more marked when the right and left halves of the lambdoid suture–the joint between the posterior margin of the two parietal bones and the occipital–are investigated.

The situation of corresponding parts of this suture are represented on the right and left halves of the recovered parts of the Piltdown skull; Dr. Smith Woodward has already.recognised the presence of those two parts of the lambdoid suture. They are indicated on the three accompanying

figures as A, B. Now on the left side of the s kull the lambdoid suture cuts the lateral line 50 mm. from the mid-line; on the right side it falls 20 mm. short of the lateral line, to make the two sides approximately symmetrical, the right lambdoid suture has to be moved outwards until it occupies the position A’ B’, shown in Fig. 1. That degree of error exceeds even the amount found in human skulls deformed artificially or deformed by disease; and points to an error in reconstruction. It will be also seen that the right and left halves of the suture, as indicated on the brain cast–the discrepancy is even more marked on the reconstruction of the skull –-have a different inclination to the mid-line of the reconstruction. It may be thought that all that is necessary to obtain symmetry is to move the parts of the right half outwards until the right and left halves of the brain cast are approximately equal in size; when this is done, it will be found that marked symmetry of another kind is introduced. In moving one part, all the other parts of the skull are thrown out of place.; the task has to be recommenced from the first initial step.

In Fig. 2 I give a drawing of the brain cast obtained when the parts of the skull are placed together according to their structural markings. There can be no doubt as to the middle line of the occipital bone; on its outer surface is clearly seen the ridge which indicates the division between the right and left halves of the neck. We may presume in this primitive man that the neck was symmetrical. The next point which has to be fixed definitely is the middle line on the roof of the skull. At first I accepted th.e middle line as fixed by Dr. Smith Woodward–an elevation on the outer aspect of the left parietal bone–near its hinder upper angle–corresponding to a wide depression which is to be seen on the inner aspect of that part of the bone. I found it impossible to obtain even an approximate symmetry of the right and left halves of the skull in all my attempts at reconstruction when I proceeded on this bais.

On comparing, the corresponding regions of the Piltdown and Neanderthal brain casts, it became quite apparent that the markings of the middle line–which I had accepted from Dr. Smith Woodward–did not represent the middle line, but a region well to the left of that line. The excavation or groove which we had regarded as caused by the longitudinal blood sinus–a channel passing along the roof of the skull under the middle line–was not due to that structure, but to the well-marked elevations of the brain on each side of the longitudinal sinus. These cerebral elevations are clearly marked in the brain casts of Neanderthal man. In the skulls of all the higher primates, the longitudinal sinus, near the hinder end of the adjacent margins of the right and left parietal bones, is marked by a narrow deep groove with dis[199]tinct edges; on the margin of the upper angle of the Piltdown fragment the edge or margin of this groove can be clearly recognised.

In Dr. Smith Woodward's reconstruction, therefore, it is not only necessary to move the fragments of the right side outwards; the left parietal bone has also to be moved outwards, or rather tilted upwards and outwards until it assumes a more vertical position, with the marking of the sinus in the middle line. When that is done, and the other parts correctly adjusted, the brain cast assumes the form and size represented in Fig. 2. I made many experiments to test other possible suppositions, but only when the fragments were placed as in Fig. 2 could I secure symmetry, and at the same time obtain all the anatomical markings in their normal situations. The brain cast obtained from this reconstruction displaces just over 1500 cubic centimetres of water. Dr. Smith Woodward estimated his brain cast provisionally at 1070 c.c.; the replicas of the brain cast which were distributed displace 1200 c.c. of water; even if the reconstruction carried out by Dr. Smith Woodward is accepted, and the right half is made approximately symmetrical with the left, the brain of Piltdown man will be about 200 c.c. above his original estimate.

In my reconstruction two other peculiar features of the original brain cast have disappeared. One is the sharp bending inwards or kinking of the temporal lobe of the brain; the other is the position of the foramen magnum–the opening in the base of the skull for the exit of the spinal cord. In the original reconstruction the lower margin of the occipital bone was brought forwards so far in the base of the skull that when a palate was articulated there was no room left for the soft palate and pharynx. The corresponding basal parts of the brain cast are, of course, also abbreviated.

I do not attach any high importance to actual brain mass; it is merely a rough indication of mental power when applied to human brains. So far as concerns the description of the actual markings of the Piltdown brain cast given by my friend Prof. Elliot Smith, I am in complete agreement, but so far as concerns general mass and conformation, it is clear, from his letter in NATURE, October 2, p- 13-1, that I am at complete variance. How far I am right – to what extent I have made an error–remains to be seen; but the publication of these drawings and observations will show that I have made every endeavour to arrive as near the truth as is possible for me.

Nature October 30, 1913

G. Ellliot Smith

[267] Now that my friend Prof. Keith has explained (Nature, October 16, pp. 197-99) so lucidly his reasons for making a big brain—case of the Piltdown fragments, it is possible to define precisely the point at issue between us.

I should say at the outset that any anatomist, working with the plaster casts but without reference to the actual fragments from which they were moulded, might solve the extraordinarily difficult problem of reconstruction of the cranium in the way Prof. Keith has explained so plausibly. But the bones themselves present features which make such a solution altogether inadmissible. Anyone who examines the left parietal and temporal bones cannot fail to recognise that there is no room for any doubt as to the relative positions of these bones the one to the other, which is not that claimed for them by Prof. Keith.

The right parietal fragment and the occipital can be put into their proper positions and the symmetry of the two branches of the lambdoid suture be restored without producing "any marked asymmetry of another kind," such as troubled Prof. Keith, and without the necessity of making any such liberal additions to the capacity of the cranium as he demands (see his Fig. 2).

The "marked asymmetry of another kind" that he could overcome only by the adoption of the most drastic measures was created wholly by his refusal to admit the possibility that the middle line in the parietal region, as determined by Dr. Smith Woodward, was a close approximation to the truth.

The determination of the precise location of the middle line in the frontal and parietal regions is one of quite exceptional difficulty, but a number of finds and considerations make it certain that it is not where Prof. Keith would place it.

The crux of our difference, then, is the criteria which Prof. Keith uses for determining the middle line in the posterior parietal region. He writes (op. cit., p. 198 et seq.): –"In the skulls of all the higher primates, the longitudinal sinus, near the hinder end of the adjacent margins of the right and left parietal bones, is marked by a narrow deep groove with distinct edges; on the margin of the right angle of the Piltdown fragment the edge or margin of this groove can be clearly recognised."

It must be remembered that the area in question (the "upper angle" of the quotation) is immediately above the middle part of the lambdoid suture, which is preserved upon the larger parietal fragment. Prof. Keith does not seem to have realised this fact, for he represents the lambdoid suture (in his Fig. 2) as a large arch (A, B, A, B) crossing the middle line a short distance below the larger bone fragment. If a series of human and simian cranial casts be examined it will be found that, contrary to Prof. Keith's statement, in a considerable proportion of them there is no trace whatever (in the place just above the lambda corresponding to that preserved in the Piltdown specimen) of "the narrow deep groove with distinct edges" on which Prof. Keith relies as his guide for the determination of the middle line. This is especially the case in the casts of the more primitive human and the simian crania, as Profs. Boule and Anthony have pointed out in their discussion of the Chapelle-aux-Saints and La Quina brain-casts.

On these grounds Prof. Keith "moved the left parietal bone outwards or rather tilted [it] upwards and outwards until it assumes a more vertical position" (p. 199). But in order to do this he had to get rid of one of "the peculiar features of the original brain-cast–the sharp bending inwards or kinking of the temporal lobe of the brain" (p. 199). If Prof. Keith had not opened out the angle between the left temporal and parietal bones the aperture of the ear would have been made to look towards the neck, when he "tilted the left parietal upwards and outwards"! But the precise relationship of the left temporal and parietal bones is not a matter of argu[268[ment but of fact; no one who examines the actual fragments and sees how precisely the edges of these bones fit one on to the other can refuse to admit that the parieto-temporal angle of Dr. Smith Woodward's restoration is a genuine peculiarity of this skull. If this is admitted it becomes impossible to tilt the upper margin of the parietal upwards and outwards. In other words, this peculiar articulation of the temporal bone afforms confirmatory evidence of the proper location of the middle line.

It is a very interesting fact that the curious conformation of the temporal region of the brain, to the reality of which Prof. Keith objects, is quite analogous to that exhibited in the remarkable cranial cast of the Gibraltar skull, of which he is the custodian, and in some of the casts of primitive crania (negro, Australian, and Tasmanian) which he kindly obtained for me.

The greater part of Prof. Keith's letter deals with the lack of symmetry in the original reconstruciton, which was due to a slight error in the positions assigned to the occipital and right parietal fragments. The need for this correction was realised before the meeting of the Geological Society last December; and this was taken into consideration when I was writing my preliminary note.

Nature November 6, 1913

Arthur Keith

You have already (Nature , October 16, p. 197) permitted me, by the use of a diagram, to demonstrate the errors in the original reconstruction; I also availed myself of that opportunity to show diagramatically the only reconstruction which gives an approximate symmetry to the right and left sides of the head, and, at the same time, places the parts in their proper anatomical positions. It is clear, from his letter (Nature , October 30, p. 267) that Prof. Elliot Smith knows of another method, one which fulfils the same conditions, but gives a much smaller brain-capacity. All that is necessary to convince me that he is right and I am wrong is a drawing of that reconstruction: one comparable with the drawings in my previous letter. I have articulated the fragments in the manner suggested in his letter, and find that the degree of asymmetry in his suggested reconstruction is as great as in the original. It is possible that I have misinterpreted some of the indications given in his letter. Any error of this kind would be cleared up by a drawing.

Nature November 13, 1913

G. Elliot Smith

[318] In my previous letters (Nature, October 2, p. 131, and October 30, p. 267) I refrained from entering into a detailed consideration of the reconstruction of the Piltdown skull, because I am preparing for presentation to one of the learned societies a full statement of all the facts and considerations bearing upon the points at issue. But I am glad to accede to Prof. Keith's invitation (Nature, November 6, p. 292) to publish a drawing of the brain cast for comparison with his (Nature, October 16, p. 198, Fig. 2).

It is a pleasure to express my hearty agreement with his appreciation of the excellence of Mr. Barlow's workmanship and of Dr. Smith Woodward's courtesy in permitting anatomists freely to handle and examine the precious fragments. Mr. Barlow's casts of the fossil bones are certainly the best examples of such modelling that I have ever seen; and I strongly resent the interpretation (op. cit., p. 292) put upon my remarks in reference to them. But even such realistically perfect copies cannot display structural details such as the texture of bone, the precise location of certain faintly marked sutures, and the nature of sutural edges of the bones; and all of these points are of crucial importance in this discussion.

On the actual fragments, for example, one can see quite plainly a part of the right half of the coronal suture (not visible on the cast), meeting the more obvious left half at an angle which must, of course be upon (or very close to) the median plane. Now this point lies upon the forward extension of the plane mm (see fig.) which was determined from other evidence (see Nature, October 30, p. 267).

Then again the texture of the bone covering the area on the brain cast near the line mm just above the point e (see fig.) is characteristic of that which comes into contact with the median longitudinal sinus. This is further confirmation of the accuracy of the determination of the line mm, namely the supralambdoid flattening, the arrangement and medial relations of the meningeal grooves, and the median groove in the frontal region, which confirm this identification of the line mm as a close approximation to the real median plane.

On these grounds the orientation of the left parietal (P) to the median plane (mm ) is settled; but we have still to determine its position in relation to the occipital upon that plane.

In spite of the extreme asymmetry of the posterior poles of the cerebral hemispheres (O and O'), the two halves of the cerebellum (Ce.l. and Ce.r.) and the lateral sinuses (L. and R.), the orientation of the occipital fragment upon the median plane is fixed as Prof. Keith has explained (Nature, October 16, p. 198).

The broken piece (b ) fits accurately upon the main fragment (O'), and as it bears upon its external face and lateral edge traces of the right part of the lambdoid suture, it is important as giving some indication of the breadth of the occipital bone at this level. [To avoid the addition of another diagram, I have inserted alongside the letter b a stippled design to suggest, in a purely diagrammatic manner, the extent and complexity of a small fragment of the lambdoid suture preserved upon the external face of the bone that covered the area b .)

Now that the occipital and left parietal fragments have been orientated upon the line mm , the problem remains of determining their relative heights the one to the other upon that line.

The left lateral sinus left its imprint upon the occipital (L.) and also upon the lower corner of the left parietal (at d ). Although the sinus is sometimes distinctly arched upward as it passes from the occipital to the parietal, the points d and the upper margin of L are as a rule on approximately the same horizontal plane, both in man and the anthropoid apes. Thus we cannot go far wrong if we bring the occipital and the left parietal into the positions shown in the diagram.

But Prof. Keith will object (Nature, October 16, [319] p. 198) that this will not bring the two halves of the lambdoid suture (cd and ab ) into symmetrical positions. To answer this criticism it may be said that the lambdoid suture in this restoration is as nearly symmetrical as it is in many ancient and modern skulls. Moreover, in the case under consideration there is the most positive evidence of a lack of complete symmetry. Not only is there the most striking asymmetry in the whole occipital area (compare O and O' and Ce.l. and Ce.r.), but the remains of the lambdoid suture itself present a marked contrast on the two sides, being quite simple on the left (cd ), but complex and dentate on the right (b ). To base any far-reaching conclusions upon the position and direction of an isolated centimetre (b ‚ of the lambdoid suture (se Nature, October 16, p. 198) is simply courting disaster. For every anatomist knows that the lambdoid is the most variable and tortuous of all the cranial sutures.

Another indication of asymmetry of the lambdoid suture is the direction of the fragment marked e . My critics may say that as it points towards the piece cd and not towards b and f , it clearly belongs to the left and not to the right half of the suture, and that it would fall into its proper position if the left parietal were moved wholly to the left side of the line mm . But such a deviation as e is quite common. A precisely similar thing occurs in the Gibraltar skull, and in the La Quina skull there is a Wormian bone near the corresponding spot on the right side.

So far I have said nothing of the right parietal fragment (P'). It bears only a very small fragment (a ) of the lambdoid suture, which, of course, must lie somewhere near the line joining e and f. Its lower margin does not quite reach the lateral sinus at f . With these and other guides (supplied by the impressions of the brain and meningeal vessels) this fragment may be orientated in a position approximately symmetrical to the left side. Incidentally, as the point a must be in the neighbourhood of the sutural line on b , the position of the right parietal fragment (P') so determined checks the accuracy of the position of the left parietal (P).

No exact symmetry between P and P' is attainable because the brain itself is not symmetrical. In the human brain the type of occipital asymmetry seen in this case (O and O') is usually associated with a greater prominence of the right parietal eminence (P'). This was the case in the Piltdown brain. In further confirmation of the reality of this it is found that the right parietal bone is very much thinner than the left, so that, as in the occipital region, the full extent of the cerebral lack of symmetry is not displayed in the outline of the skull.

In making the drawing illustrating this letter I have used a cranial cast which Dr. Smith Woodward kindly sent me a few weeks ago, but have made some slight alterations in the positions of the two parietal fragments.

In conclusion I should like to say how much I am indebted to Prof. Keith for all the help he has given me in my investigations, not only by allowing me to make use of all the valuable material in the museum of the Royal College of Surgeons, but also by discussing with me frankly and openly all the points in dispute concerning the Piltdown skull itself. In the earlier part of July, working with the cranial casts, he seemed to me to have established a good case for his mode of reconstruction; but from the moment I began to examine the actual fragments (August 13, 1913, the day after the discussion of the matter at the International Medical Congress) I became convinced that his solution of the problem was an impossible one. It was this personal experience of the importance of working with the real things that I had in mind when I was writing my last letter (Nature, October 30, p. 267).

Nature November 20, 1913

Arthur Keith

In the accompanying tracing of Prof. Elliot Smith's reconstruction I have indicated the longitudinal blood sinus which sweeps widely (10 mm.) to the right as it passes between the occipital poles of the brain. The left pole exceeds the right to a degree which is seldom seen in even the highest forms of modern human brains. Seven years ago Prof. Elliot Smith published a short paper (Anat. Anz., 1907, vol. xxx., p. 574), which is justly regarded as authoritative. He directed attention to the preponderance of the left occipital pole of the brain, and attributed that preponderance to the specialisation of the right hand; only the slightest degree of asymmetry is observable in anthropoid apes. Indeed, at that time Prof. Elliot Smith definitely stated that he regarded symmetry of the occipital poles–in my opinion an absolutely just deduction–as a simian character. He will, therefore, if he retains the present reconstruction, have to modify to some extent the opinion he has expressed of the brain of Piltdown man–that it is "the most primitive and simian brain yet recorded." As regards the asymmetry of the occipital poles, it is, in my opinion, ultra-modern.

Prof. Elliot Smith has frankly stated that his reconstruction is not, in the strict sense of the word, a cranial cast–a cast taken from the interior of a reconstructed skull; it is a reconstruction built up–as the original brain cast must have been–from impressions taken from the inner or cerebral aspect of the cranial bones. To test such a brain reconstruction the actual fragments of the

Fig. 1.–Tracing of Prof. Elliot Smith’s reconstruction of the brain-case with certa in addition. (Half nat. size.)

skull must be placed over the corresponding parts of the brain cast. When that is done it is at once seen that in securing a symmetry of the brain hemispheres the corresponding parts of the skull are thrown somewhat out of position. On the tracing of the reconstruction (Fig. 1) I have drawn a line, x–y , across corresponding angles of the parietal bones. That of the right side is a centimetre higher than on the left; on the right side the lambdoid suture passes outside the 50 mm. vertical line; on the left it stops short of that line.

Fig. 2.– Tracing of the right fragment of the parietal (Piltdown) fragment (broken line) superimposed on the right bone (continuous line). (Half nat. size.)

I have not entered into a discussion on the markings which indicate the middle line of the skull for this reason. A very considerable experience in attempting to reconstruct ancient and modern skulls from fragments has convinced me that if a wrong bearing is taken–if one misidentifies any point in the middle line–unless it be a very slight error, that misidentification will find the reconstructor out, and his task will be brought to a halt by the development of a degree of asymmetry. If, on the other hand, points are rightly recognised–often if has to be by repeated experiment–then the parts fit easily together, provided there is a sufficiency of them, and in the case of Piltdown there is an ample sufficiency. I look upon the problem of rightly reconstructing a skull as similar to that of replacing the fragments of a broken vase of symmetrical design. Given the fragments of the greater part of one half and a part of the other, there cannot be two reconstructions. All the parts may be got together, except one fragment. The remaining fragment is evidence that the task has not been accomplished. I know very well that my friend Prof. Elliot Smith is searching for a true representation of the brain-state of the very earliest human form that can claim any relationship to modern men; I hope I may claim the same spirit for myself. I also admit that he has gone a considerable way towards what, in my opinion, must have been the original form. The points on which we disagree are now apparent, and I am content, having had an opportunity of presenting my case, to leave the final decision to the future.