One of the most controversial subjects in the whole of European history is the physical composition of the Keltic peoples. The name Keltic has been applied to many racial types, real and imagined, from short, brunet, round heads to blond brachyephals and Nordics. Many modern prehistorians take the stand that the Kelts were everywhere a small minority of aristocrats and conquerors, and that no special racial type accompanied their expansion in Europe. This position, however, becomes invalid when we examine the actual skeletons of Keltic speakers. There was a Keltic physical type, which the Kelts carried to their primary areas of colonization, and which will be described shortly.

Although earlier identifications, however likely, are still questionable, we may state that the Kelts as such first appeared in the European historical setting about the year 500 B.C. with the beginning of the La Tène civilization. The home of the Kelts, or at least the country in which they developed this brilliant Iron Age culture, lies without reasonable doubt in southwestern Germany, in the upper drainage of the Rhine,²⁶ a country which had formed the western section of the original Hallstatt area. The easternmost outposts of the early Keltic domain were Bohemia and Galicia, while, on the west and south, it touched the territory of the Ligurians and of the Rhaetians. The Kelts, therefore, were situated northwest and west of the Illyrians proper, and south of the Germans, who at the time were confined to Scandinavia and northwesternmost Germany.

The Keltic languages are very closely related to the Italic group, of which Latin was a derivative. The period in which the Keltic languages became differentiated from other forms of Indo-European speech must, therefore, be as old as the departure of the ancestors of the Italici for Italy, and therefore must lead back to the Bronze Age.²⁷ Keltic, like Italic, is divided into two branches - P-Keltic and Q-Keltic. It is considered likely that the phonetic separation which split both of the linguistic groups took place independently in each, and that the tendency for such a division was inherent in both Keltic and Italic at the time of their separation from one another. We do not know at what time the Goidelic or Q-Keltic dialect split off from the Brythonic or P dialect, but this cleavage again must have occurred at a reasonably early period, since the division was complete at the time of our earliest knowledge of these languages. Q-Keltic has survived only in Ireland, Scotland, and on the Isle of Man. All other known dialects, living and extinct, from Asia Minor to Wales, have been of the P variety.

The Keltic expansion, which began about 500 B.C., was a rapid and extensive one. The Kelts were an extremely mobile people who conquered md wandered far, and at the time of their expansion were apparently numerous as well. Their well-known migrations carried them over the Alps into Italy, down into southeastern Europe where they invaded Greece, and even over into Asia Minor where they established the short lived Galatian colony. Their main expansion, however, lay to the west. Belgium and northern France became great Keltic centers, from which some of them migrated down into northern Spain. This westward movement carried them also into the British Isles, where the Q-Keltic people settled Ireland, and their P-Keltic brethren established themselves in England and Wales. Large sections of Scotland were to remain free for the most part from these Keltic invaders until after the time of Christ, when the Goidels crossed over from Ireland.

The question as to the linguistic identity of the previous inhabitants, the Picts, is an open one. At present, the tendency is to consider them, and the pre-Goidelic Cruithni of Ireland, as speakers of some early form of Keltic. The further question as to whether or not the Goidels crossed England in their journey to Ireland is likewise open, but the prevailing tendency is to bring them over the old sea road from northern Spain, which they had previously entered by way of France, and to deny that they sojourned in England at all.

In their period of development in southwestern Germany, the relationship between the Kelts and Illyrians must have been intimate, for the Kelts received iron from a Hallstatt source, and were actually, during the Early Iron Age, participants in a Hallstatt form of culture. The major factor which served to differentiate La Tène from Hallstatt culture was the incorporation, by the former, of many elements derived from the classical Mediterranean world. The Kelts were situated at a favorable spot for the reception of such influences; Greek influences moved up the Rhóne and Sabne from Marseilles, while those from Rome crossed the Alpine passes into Bavaria and Switzerland and thence into the Keltic homeland.

In addition to the Hallstatt Iron Age base and classical accretions, we must further acknowledge the influences of some eastern European grassland culture, for the Kelts rode astride as well as in chariots, and the P-Kelts introduced trousers to western Europe. This garment was central Asiatic in origin, and was typical of the Scyths, whose period of cultural efflorescence in the east was contemporary with and parallel to that of the Kelts in the west. Philologically, there are a number of close linguistic connections between the Kelts and the Indo-Iranians, which may reflect this or an earlier cultural contact. It is most likely, however, that the principal contact between the Keltic-speaking peoples and the Iranian horsemen of the eastern European plain took place during the early years of the great Keltic expansion.

Turning back from Keltic expansions to Keltic origins, we find no cultural disturbances in southwestern Germany which would permit the arrival of the Kelts from elsewhere between the Hallstatt epoch and the early La Tène. Before the Hallstatt, however, the spread of the Late Bronze Age Lausitz culture into this region from eastern Germany may conceivably have brought a large number of people, impossible to identify because of their practice of cremation. These people may well have been the bearers of Keltic speech. Since the related Italici were themselves Urnfields cremators before they succumbed to indigenous burial rites in Italy, this identification is rendered more than likely. Hubert has, indeed, postulated an earlier Ligurian-speaking population in the Keltic cradle-area.²⁸

The derivation of the Kelts from a Hallstatt cultural horizon, in part of the earliest region of Hallstatt development, while the main current of Hallstatt cultural expansion was borne by Illyrian speakers, seems incongruous. One must remember, however, that the Nordic skeletal type with which the Illvrians were identified in Lower Austria was confined, in its purely dolichocephalic form, to the lowland country north of the Bavarian foothills, while the Keltic area of development was, in its strictest limits, within the highland zone. Here the Kelts developed their own culture independently of the Illyrians and retained their own language.

Keltic cranial material from the southwest German center of Keltic development is surprisingly scarce. Schliz has described six skulls, and notices of three others have appeared in more recent publications.²⁹ Of these nine, one is dolichocephalic, four are mesocephalic, and four are brachycephalic. Although this small group is far from sufficient to disclose the racial type of the Kelts in their homeland, it is enough to show us that a round-headed element played a considerable part in the development of this ethnic group. The brachycephals involved are large headed and powerfully built, with long faces, and rather high orbits; the foreheads are sloping and only slightly bowed at the junction of the facial and cranial planes. The inference is that these brachycephals were derived from the older combination of Bell Beaker and Borreby types which was formed in the upper Rhine country at the beginning of the age of metal, and which persisted into the Hallstatt period. These seem to have mixed with the expected intrusive Nordics. We must really wait until we examine larger series of Keltic crania from elsewhere, however, before passing judgment on the final result of this blend.

A better picture of the La Tène type may be obtained from the study of its early eastern extension. Hellich's series from Bohemia³⁰ (see Appendix I, col. 33) is the only single group of central European La Tène crania of any consequence. This includes 27 male crania, most of which are dolichocephalic, but which contain a significant minority of brachycephals. In general, the La Tène skulls are not in any important metrical way distinguishable from those of the preceding periods of which we have clear knowledge - that is, Aunjetitz and Hallstatt. They represent merely a sub-variety of the same general combination of types, with a brachycephalic accretion which makes the total series mesocephalic.³¹ But there are other features, however, which render them as a group slightly different; the vault has a tendency to be low in proportion to its breadth, and the upper face is long in proportion to the total face, for the Keltic jaw, although broad at the gonial angles, is not as deep as that of other Iron Age Nordics. A composite series of eleven male crania from the type site of La Tène on Lake Neufchatel in Switzerland, and nearby burial places,³² is almost exactly the same as the Bohemian series; the vaults of the Swiss La Tène people, who may in part be identified with the Helvetii, are even lower than those of the Bohemians. As one might expect, the Swiss series con-tains a number of high brachycephals, with cranial indices as high as 90;³³ but on the whole, most of the few Kelts whose remains have been studied in Switzerland were no different from those in Bohemia.

Less than a dozen skulls serve to identify the Keltic racial elements in Austria and in the Dinaric Alpine mountain zone.³⁴ On the whole, this evidence is not satisfactory, but it serves to indicate that the regular mesocephalic type and one or more types of brachycephals were present. The most southeasterly Keltic skull known is one from Kupinovo, near Belgrade in Serbia, which belonged to a Dinaric brachycephal similar to those found at Glasinac, and this again witnesses the persistence of this Dinaric element during the Iron Age in or near the modern Dinaric area.

Before turning to the abundant remains of the Kelts in France and the British Isles, it may be well to review what evidence we have for their racial type in central Europe. Here the Kelts seem to have been a composite people, a blend of the different brachycephalic elements left over from the Bronze Age in the mountainous zone of southern Germany, and invaders of Nordic type from the plains to the north and east. One supposes that the Keltic linguistic element came with the later group.

Sculpture from Greece and Rome gives us a picture of the living Kelts who reached the lands of classical civilization by eastward and southward movements. The well-known Dying Gaul and similar statues show a strongly muscled type with mesocephalic or brachyeephalic head form, a rather short face with a square jaw. a straight and rather prominent mesorrhine nose, with horizontal or elevated tip and full nostrils, heavy browridges, a broad forehead, and stiff, bristly hair. This type, while familiar enough in western Europe. is not one which accords with the majority of the Keltic skeletons. The typical Keltic face was long in the upper portion, shallow in the mandible, long and narrow of nose, often with a convex profile, and the forehead was extremely sloping and the vault low. This has its most frequent counterpart today in the British Isles. While the type selected by the classical sculptors to represent the Kelts must have had its living models, these may have been drawn from the brachycephalic minority.

Most of the La Tène material from France comes from the north, from the Maine region, where the Keltic settlement seems to have been particularly strong. Fortunately, large and competent series of the Gauls of this district, before and after the Roman conquest, furnish adequate information.³⁵ (See Appendix I, col. 34.) Both groups are alike, showing that submission to Roman rule did nothing to change the physical type of this particular people.

The Gauls as so represented were mesocephalic, mesoprosopic, and on the upper borders of leptorrhiny. The vault, as with all characteristic La Tène Keltic groups, is not distinguished for its height, and in the large and more reliable post-Roman series, it is definitely low. Like their relatives in central Europe, these Gauls were not noted for tall stature; a mean of 166 cm. is only moderate.

In other parts of France, the Keltic racial continuity was of variable intensity; in Lorraine and Beaune,³⁶ the usual type was found; but in Haute Savoie and Vendée the earlier brachycephalic population is strongly represented in Keltic tombs,³⁷ while out on the tip of Brittany, Neolithic sinivors of Mediterranean type, with perhaps some Gaulish admixture, persisted until the period of Roman conquest.³⁸ Only in the north, therefore, did the Kelts make a firm imprint in the early population of what was to become the French nation.

The Kelts in the British Isles are known to us by a large series of Brythonic crania from England and southern Scotland, assembled by Morant³⁹ (see Appendix I, col. 35); these are three millimeters longer headed than the Bohemian and Swiss series, but nearly identical in vault dimensions with the French; facially they are the same as all of the others. Smaller collections of Goidelic crania from Ireland show the skulls from this country to be exactly the same as those from Great Britain.⁴⁰ Several morphological features distinguish these skulls, of the typical, or mesocephalic, group - which in the British Isles seems largely to lack the brachycephalic minority which accompanies the main type in central and eastern Europe. The forehead is quite sloping; the vault, when seen from behind, gives a cylindrical impression, rather than that of a rhomboid or rectangle, as with other Nordic crania. The upper face is quite long, the mandible wide at the back, and relatively shallow. The nose is often very prominent.

The skeletal material from Ireland (see Appendix I, col. 36) is not numerous enough to permit regional studies, or other statistical niceties; it in Great Britain there are, on the contrary, a number of local series sufficient to show that the racial complexion of that island was not, during the Iron Age, completely uniform. One of these, that of the erroneously named "Danes' Graves" at Driffield, Yorkshire,⁴¹ containing 29 male crania, is identical in every known respect with the Aunjetitz skulls from central Europe - a pure (if the adjective pure may be used of a composite type) Hallstatt or Nordic local population; purely dolichocephalic, in contrast to the usual Keltic mesocephaly; and relatively high-vaulted, again non-Keltic, although the stature, 167 cm., is presumably no different from that of the Kelts.⁴²

It is impossible to derive this group from the local Neolithic, which was noted for its extreme absolute cranial length; or from the dolichocephalic element of the Bronze Age, which was again larger, longer, and higher-skulled; it resembles not only the earlier Aunjetitz and Hallstatt, but also, although to a lesser degree, the contemporary Scandinavian Iron Age people in the period immediately before the Germanic Völkerwanderung. All of the archaeological material found in the Danes' Graves has never been satisfactorily identified.⁴³ Although the dominant Keltic tribe of that neighborhood, the Parisii, seems culturally represented, it is unlikely on archaeological as well as on racial grounds that the majority of the men buried in these graves came from the Maine, whence the usual Brythonic tribes migrated to England. Two of the fibulae found in the scanty remains have Scandinavian affinities; despite this clue, however, we must leave open the question of the immediate origin of the Danes' Graves people, and render the verdict: "Central European Nordics found in Yorkshire during the late Iron Age, provenience unknown."

Another local group which shows aberrant tendencies is that of eleven male crania from Berkshire, of which the length, breadth, and circumference alone are available;⁴⁴ the figures are 193.3 mm., 149.6 mm., and 552.2 mm. The cranial index is 77. These mesocephalic crania are so much larger than those of the total Iron Age population that some other origin must be postulated. One recalls the extravagant dimensions of both Neolithic and Bronze Age crania in England, and may only suppose that this local group represents a relatively unaffected survival. Since both Bronze Age and Neolithic racial types may be picked out of any moderate-sized gathering of living Englishmen, or of their transatlantic relatives, it is not surprising to find a few in Berkshire during the Iron Age.

The descriptions of the Kelts, in Britain, in France, and in other parts of Europe, at the hands of classical authors, give us a definite picture of their pigmentation. Blondism was by no means characteristic of the Kelts as a whole. Rufosity was common, and the hair color was essentially mixed. Caesar himself noted the contrast between the ordinary Gauls and the partly Germanic Belgae, to whom he had to turn to find real blonds for his triumph. Furthermore, the Romans noted the Keltic practice of bleaching the hair to simulate a blond ideal, as in Greece.

On the whole, the Kelts were a mixed group in race as in culture; their ancestry includes both long heads of some central European Nordic type, which was in turn a combination of several Mediterranean sub-types, and brachycephals from the region in southwestern Germany in which the Dinarics of Early Bronze Age introduction had blended with earlier round heads of Mesolithic origin. Out of this combination, the Kelts developed an easily identified national type, of considerable constancy, which was a to be of some importance in the world, especially in Britain and the nations derived from her.

Notes:

26 Hubert, H., The Rise of the Cells, p. 147.

27 Although one school of Italic scholars derives the P-Italici from north of the Alps in Iron Age times, all admit the Bronze Age dating of the Q-Italic arrival. For the details of this controversy, see Whatmough, J., The Foundations of Roman Italy.

28 Hubert, H., The Rise of the Kelts, p. 159.

29 Jacob, C., AFA, vol. 20, 1891-92, p. 181.
Ortmann, R., JVST, vol. 15, 1927, pp. 56-59.
Schliz, A., AFA, vol. 37, 1910, pp. 246-251.

30 Hellich, B., Praehistoricke lebky v Cechách ze Sbírky Musea Království Ceského.

31 Schliz's series of 14 crania from Bohemia, 3 from Moravia, and 2 from Silesia do not differ from those measured by Hellich. Schliz, A., AFA, vol. 37, 1901, pp. 246-251.

32 Virchow, R., ZFE, vol. 16, 1884, pp. 168-181; ibid., vol. 18, 1886, pp. 561-566.
Lagotala, H., BMSA, ser 7, vol. 3, 1923, pp. 4-9.

33 Schlaginhaufen, O., AFSA, N. F. Bd. 38, 1936, pp. 226-236.

34 Pöch, H., MAGW, vol. 56, 1926, pp. 255-270.
Lebzelter, V., WPZ, vol. 22, 1935, pp. 104-105.
Luschan, F. von, MAGW, vol. 8, 1879, pp. 85-89.
Schliz, A., loc. cit.

35 Raymond, P., RP, vol. 2, 1907, pp. 10-22, includes 20 males.
Wallis, Mrs. Ruth Sawtell, unpublished measurements in Musée Broca, Ecole d'Anthropologie, and Musée d'Histoire Naturelle. Includes 28 pre-Romans and 83 Gallo-Romans, all males.

36 Hamy, E. T., Anth, vol. 17, 1906, pp. 1-25; vol. 18, 1907, pp. 127-139.

37 Baudoin, Marcel, BSAP, vol. 6, 1912, pp. 321-346.

38 Vallois, H. V., Les Ossements Bretons de Kerné, Toul-Bras, et Port-Bara.

39 Morant, G. M., Biometrika, vol. 18, 1926, pp. 56-88. Also Hook;
Beatrix, and Morant, G. M., Biometrika, vol. 18, 1926, pp. 99-104.

40 Martin, C. P., Prehistoric Man in Ireland. Twelve Iron Age skulls are listed.

41 Wright, W., JRAI, vol. 33, 1903, pp. 66-73; Archaeologia, vol. 60, 1906, Pt. I, pp. 313-324.
Mortimer, J. R, Man, vol. 9, 1909, pp. 35-36.

42 We know the stature of Kelts in the British Isles only from a small Irish group, and by inference from comparison with mediaeval English counterparts of Iron Age skeletons.

43 Greenwell, W., Archaeologia, vol. 60, part 1, pp. 251-312. Bremer, W., Real, vol. 1, pp. 229-230, article "Arras."