Next to stature, which is of interest to many others besides anthropologists, our data are fullest on the cephalic index, for this ratio has been the favorite of both professional and amateur students of race ever since its invention by Retzius in 1842. The same remarks on the method of plotting the stature map apply to that of the cephalic index (Map 6). Here the only region of comparative uncertainty lies in the southeastern corner, in Iran, where some rather extensive boundary stretching has been practiced.

The distribution of the cephalic index within the area covered by this map is a complex affair, and cannot be interpreted hastily. Many factors and many events have contributed to this state of complexity, which the map only partly represents. One must remember that, as in the stature map, the scattered bands and villages of Lapps have been schematically united into a nucleus in northern Scandinavia, Finland, and the Kola Peninsula. Furthermore minority groups such as Jews, Gypsies, and others, have been omitted, since in no region large enough for schematic representation are they found in a majority.

The most striking feature of the map, and in fact, almost its only uniformity, is the steady band of almost pure dolichocephaly which extends south of the Mediterranean from the Atlantic coast of Morocco across North Africa, Egypt, Arabia, and Persia into Afghanistan; to continue. off the map, over Khyber Pass and into the Indus Valley. This band represents the greater Mediterranean race in its post-Pleistocene homeland. Small spots of mesocephaly in the Moroccan mountains, in Kabylia and in the Aures, and along the Tunisian coast, show the relatively restricted zones of survival of earlier Mediterranean mesocephals and, to a lesser extent, of Pleistocene North African men; except for the Tunisian coastal centers, where the strong concentration of Punic and European populations in pre-Arab times is no doubt partly responsible.

The extreme long heads, concentrated in the Hoggar and in parts of the Algerian plateau, are the Tuareg and the purer families of ancestral nomadic Berbers, preserving the head form which they brought from East Africa, their Hamitic homeland. The heavily dotted stipple represents Mediterraneans of Neolithic age and Arabs, with an infusion of the Hamites, while the light dotting represents more clearly the Hamites themselves. This is a distinction which should not be pressed too far, but which may still be made, for the lightest stippling is found in nomadic Berber strongholds.

Farther east the desert tribes of Libya, and the oasis people of Siwa, are extremely long-headed, in a truly Hamitic fashion; the inhabitants of Sinai, and some of the tribes in the Nejd, as some of the Mesopotamian Bedawin, and groups in Iran, fall into the same category. Here in the east we approach the zone of hooked-nosed long heads, quite different in facial form from the Hamitic increment farther west. Around the Persian Gulf is a ring of higher indices, representing a maritime population which we shall encounter later in the coastlands of southern Arabia, off the present map. The long headedness of inland Arabs, whether nomadic or agricultural, continues without a break south of the present map into Yemen and to the northern and western borders of the Ruba' el Khali.

In Europe itself, long-headed total populations are rare. Only in parts of Portugal, in fact, are regional indices under 76 to be found at all. Europe on the whole is a brachycephalic or mesocephalic continent. Mean indices between 76 and 79, belonging to high dolichocephals and low mesocephals with brachycephals in the minority, are found in a few places. One, the most continuous area, lies in the northwest; it includes the British Isles, most of Holland, parts of Belgium, and the Palatinate - old Frankish country - and most of the Scandinavian Peninsula, along with the coastal lands of Finland, and with Esthonia and Latvia.

The regions just enumerated may be considered in a way a unit; most authorities would call this, as with stature, the Nordic racial territory, and so it is in the accepted sense. Another belt is that of the Iberian Peninsula, the Dordogne Valley in France, Sardinia, Corsica, the Balearics, the toe of Italy, and Crete. To this may perhaps be added part of the corresponding area in the British Isles, and parts of the eastern site of the Balkan Peninsula. This is what remains of the brunet Mediterranean race per se in Europe; isolated island groups, a peninsula which throughout history has been more African than European, and remnants of the old Mediterranean bloc of the shores of the Black Sea and the Aegean.

Where, we ask, are the descendants of the Danubians, the Aunjetitz Nordics, and their Iron Age successors in eastern and central Europe? Only in the mesocephalic belt across eastern central Russia, and the region immediately north of the Caucasus, and again in the central and eastern Balkans, do traces of the original head form of these peoples appear, emerging as that of a population bound to the soil. Perhaps in the tall stature and high mesocephaly of the Don country there is also some trace of the Scythians. The country between the northern shores of the Caspian and the middle Baltic does indeed form a zone of relative long headedness between the mongoloid brachycephaly of central Asia and the European brachycephaly of central Europe.

This central European brachycephaly may not be treated as a completely unified entity. In the first place, we find its westernmost nucleus in southern France in the Massif Central, which is the home of the Alpine race in its truest form. Here extreme round headedness such as is seldom exceeded elsewhere in the world is located. The valley of the Rhóne forms a partial gap, beyond which lies another brachycephalic zone in eastern France, especially in Burgundy and the Jura, and adjacent portions of Belgium. Here again we find a high zone of brachycephaly, accompanied, as we have seen, with a greater stature than that found in the western Alps, and as we shall see later, a lighter pigmentation. Here is another brachycephalic nucleus representing a different racial concretion from that first mentioned. One observes that in the upper Rhine Valley and in northwestern Switzerland, as in Lower Austria, this zone of extreme brachycephaly is broken, while a northern colony of it is found in Bavaria, Bohemia, and Silesia.

In the Tyrol, southeastern Switzerland, and most of northern Italy is another nucleus, which is the home of the western branch of the Dinaric group, associated largely with the center of Rhaeto-Roman speech. These linguistic fossils are survivors of the pre-Germanic population of this region. Most of Austria itself runs longer headed, owing, no doubt, to the strong concentration of Germanic peoples there. The Dinaric region proper, extending from Bosnia to southern Albania, follows the mountain range, which in turn lies close to the Adriatic coast. The center of highest brachycephaly lies in southern Albania, in the Tosc country, well south of the center of tallest stature. The southern brachycephalic zone, of which it is the nucleus, extends far into Greece, along the western coast, from Epirus to the Gulf of Corinth.

The curve of the Carpathians forms a brachycephalic barrier, within which all peoples represented, except for the Hungarian Szeklers, are very round headed. This infra-Carpathian brachycephaly pervades all other groups regardless of language, culture, or history. Beyond it lies the relatively long-headed expanse of the Polish, Ukrainian, and Moldavian plain.

As we turn to Asia Minor we see other instances of extreme regional brachycephaly. The Armenians, some of the Syrians, especially the Alouites, Lebanese, and Druses, are the roundest headed of all in this region. The Anatolian Turks, being typically brachycephalic, in this respect resemble modern representatives of the pre-Turkish peoples, of this region, notably the Armenians.

The cephalic index map, like that of stature, shows that the Mediterranean Sea is by no means a racial unit. Some of the lowest and some of the highest cephalic indices in the world are found in close proximity to its shores. Another notable lack of continuity is seen in the far north. The hunting and fishing peoples, so consistently short of stature, are very variable in head form. The Lapps alone are consistently and extremely brachycepjialic. The original mesocephalic head form typical of the Finns in their native habitat may still be observed in the regions occupied by Finnish survivals in central and northern Russia.

On the whole, the distribution of the cephalic index in Europe and adjacent countries is extremely significant when one remembers the historical and archaeological background, but viewing its present distribution alone one might easily form numerous false ideas about racial origins and continuities. It is sufficiently clear, however, that the zone of extreme brachycephaly in central Europe has several nuclei, and is separate from the Anatolian-Caucasic center and from that of the mongoloids of central Asia.

One last factor remains to be mentioned, and this is the ultra-peripheral distribution of moderately high cephalic indices on the very westernmost fringe of Europe. One notices that southwestern Ireland has a mean cephalic index of 80 or over. Little spots of this same condition occur in northern Scotland, the Shetlands, the West Frisian island chain, in Fehmarn, and in points along the western Norwegian coast. This hypermarginal brachycephaly is peripheral to the dolichocephaly of northwestern Europe, which in its turn is a survival. The suggestion is that this round-headed tendency of the extreme western fringe is in the nature of a Palaeolithic reemergence. The third map of this series (Map 7), is intended to show the distribution of absolute head size. Head size ideally should be a measure of the cubic capacity of the cranium, and capacity may be estimated upon the living by the use of the three dimensions, head length, head breadth, and auricular head height. Unfortunately, however, as already explained,¹⁷ auricular head height is for the most part an unreliable measurement,

and it would not be possible to construct a map covering a large area in which this was a component dimension. For this reason head size is here expressed simply by the sum of the length and breadth in each sample used. It so happens that large heads in the length-breadth sense are frequently high heads as well, so that there is little chance that the omission of the height dimension has falsified the appearance of head size conditions.

Head size is, in the first place, wholly unrelated to head form. Some of the largest heads are found among both dolichocephals and brachycephals, and the same is true of some of the smallest heads. It seems, however, to be closely correlated with total bodily bulk, and hence with weight, although not with stature.¹⁸ This principle applies to other animals as well as to man. Brain size is, after all, a component element of bodily bulk, and the requirements of the organism in the matter of nerve tissue depend apparently upon total size rather than upon the relative degree of attenuation of extremities. We have seen that cranial size is an important racial diagnostic in the cranium, and there is every evidence that it is equally important in the living.

The map which shows the distribution of this trait is not, however, as reliable as the two which precede it. Lacunae have been filled in accordance with general racial trends and by the conversion of modern cranial material to living standards by fixed additions to allow for the soft parts.¹⁹ The areas which are least reliable are Portugal, Spain, much of France, and portions of western Germany. The Balearics and Sicily were filled in by inference. However, the data are sufficient to assure us that the general picture is correct, although the boundaries may well be inaccurate. The map will serve our purpose, and cannot lead us far, astray, if we do not lean too heavily on it, or follow it in too much detail.

The first impression that the map gives is one of a concentric distribution of head size with Germany, Belgium, and northern France as the focus of greatest volume. From this focus bands of diminishing size stretch like bars dexter to the Persian Gulf. This pattern is broken in the Middle East by the intrusion of relatively large-headed mongoloid peoples from central Asia, and of non-mongoloid dolichocephalic Turkomans, Azerbaijanis, and Kurds.

Studied in greater detail, where detail is justified, this basic pattern does not break down, but other facts appear. In the first place, Ireland as a whole has the largest heads of any country excepting Belgium. A vertical line divides Ireland into a western, and especially southwestern half, with heads as large as the largest elsewhere, and an eastern, and especially northeastern, half with heads which although smaller, are still large by European standards. Iceland again is an area of maximum head size, and so are the Shetland Islands. Small regions of large head size appear along the Norwegian coast. The regions mentioned in this paragraph undoubtedly represent the maximum survival of Pleistocene European man of the Brünn race in the northwestern portion of the continent. They coincide to a certain extent with the hypermarginal distribution of high mesocephaly and low brachycephaly.

But there remains the bloc of large heads running from the Seine to East Prussia, and concentrated in Belgium and in the lower Elbe country. Here large heads are associated with brachycephaly, of varying degrees, but usually of a moderate order. This region has a much larger-headed population than has most of Sweden and Norway, and most of England and Lowland Scotland. The brachycephals of this large continental bloc all have head lengths which elsewhere go with dolichocephaly. The Fehmarn islanders, for example, whose small home is just south of the Danish Archipelago, have a mean head length of 193.5 mm., and a cephalic index of 83.6.²⁰ Their head breadth of 161.8 mm. is tremendous. In our historical chapters, we encountered but one racial type which consistently presented the combination of brachycephaly with great head lengths. That was the type found at Afalou and Ofnet, and in the Danish middens, and which was given the name Borreby. As will be seen later, the Borreby race has reemerged in the country where it was located during the Mesolithic and Neolithic periods, and it has become the most important single racial element in modern Germany.

Palaeolithic and perhaps Corded survivals are to be seen in the large heads of the Finnish coast and northeastern Sweden; the track of German colonists in late mediaeval times is evident in Hungary and Rumania. The Basques have heads of considerable size also, and there seems to be a significant nucleus of large heads in the Dordogne, where, as will be seen later, a long-headed, brunet Upper Palaeolithic survival seems indicated, as in west-central Wales.

The zone of moderate head size lying between Germany and Poland on the one hand, and eastern Russia and the Caucasus on the other, seems to reflect an earlier Danubian and Nordic condition. In North Africa and southern Italy, small or medium-sized heads seem marginal and go with the older Neolithic Mediterranean element. The Hamites brought larger heads, such as are to be found today among Galla,²¹ and among other predominantly Hamitic peoples.

The tendency of the Hamites to large head size has divided the erstwhile unified Mediterranean racial zone, which stretches across the whole lower quarter of the map, into a western and an eastern compartment. The eastern sector, from Cyrenaica to India, shows the small head size which apparently formed a cranial interlude in North African history between the end of the Capsian and the Hamitic invasions. As one leaves the map and passes into southern Arabia and Baluchistan, the heads grow smaller than any here designated. Here total length-breadth combinations of 328 mm. are found in the Hadhramaut and among Brahui. This zone which stretches along the northwestern shore of the Indian Ocean is part of the so-called Veddoid racial area, which does not extend into Europe or any region nearly approaching it. The racial character of the people inhabiting this zone can best be described in a more detailed chapter to follow.

One of the most important results of the plotting of the head-size map is the discovery that the brachycephals of the white race and of Europe are not at all a unit in this respect, since they follow general racial zones which have no reference to head form. One may divide them into several sub-groups on the basis of head size alone. The Lapps, who in their pure form are hyperbrachycephalic, have very small heads. The other brachycephals of northern Europe, those concentrated in Germany, southern Denmark, Belgium, and France, form the largest-headed group. These may be considered, tentatively at least, of Borreby derivation or inspiration. The Alpines of the Massif Central in France separate themselves clearly from this nucleus, with an emphasis on moderate head size. Although the regional data in France is poor, in this case it is sufficient to warrant the present conclusion. The Dinarics are also moderate in head size, despite the coincidence of taller stature; only the Montenegrins themselves and the Albanians north of the Drin have truly large heads. The extreme hyperbrachycephals of southern Albania and Epirus are again of medium head size, like the Central French Alpines. The brachycephals of the Hungarian plain, and of the Carpathians, are for the most part also moderate.

When we leave Europe and move to western Asia, we find that the Asiatic Dinarics and the so-called Armenoids are in some areas smaller headed than the European Dinarics; the Armenians themselves have heads approaching Dinaric standards, but they vary regionally, with the largest heads in the northeast, toward the Caucasus. The brachycephalic Turks of Asia Minor are actually small headed, as are most of the Syrian brachycephals and the Iranian-speaking round heads of the Pamirs. The fringe of round heads along the southern Arabian, Persian, and Baluchistan coasts are very small headed, in a quite un-European sense.

What are we to make of all this? The answer cannot be given as yet in final form, but several suggestions present themselves.

(1) Head size, being a correlate of gross bulk, seems in general to be associated with regions of relative chill and humidity, all else being equal. The water content of the human body is greater where evaporation is east. In this way the flaccid Teutons and the fog-wreathed Irish in their moors and bogs have the heaviest bodies and the largest heads, while the indigo-stained Arabs, living on the utmost margin of desiccation, reach the opposite extreme in liquid economy. Man is not a water-storing creature, like the cactus and the camel.

(2) The largest-headed peoples are unreduced survivors or counterparts of Upper Palaeolithic man, who was a large-headed and presumably large-bodied animal. This applies both to dolichocephals and brachycephals. Brachycephaly is a mutative incident which may occur in any region or race, and head size may be more important than head form as an indication of ultimate genetic derivation, again all else being equal.

It seems to me that somewhere between these two hypotheses lies the truth. Environment, which in the last analysis controls body size, must also eventually control the bulk of the head. But at the same time, genetic tendencies to absolute head size are inheritable, and without regard to head form. Hence early racial connections, under equal environmental conditions, may be better revealed by the size than by the shape of the vault. The heads of some people have remained constant in size and form; others have been reduced, brachycephalized, or both. But brachycephalization may take place without reference to body size, while reduction in head size is a corollary of general reduction. Here, as in general, the explanation of a given head size is an historical matter.

Other criteria of the head and face would be difficult to plot. Face size, in general, is larger among the larger-headed and taller peoples of the northwest, and among those of mongoloid affinity in the east. Most branches of the Mediterranean stock proper are characterized by relatively short and relatively narrow faces. The zone of long heads from Morocco to India is also a zone of small faces. This smallness, however, has as a rule no reference to the nose, which is one of the best racial criteria which we have, and one which is extremely significant. Unfortunately accurate charts cannot be made, since technical discrepancies render the use of statistics based on this organ almost useless in a large compilation.

The nasal index among European peoples is typically leptorrhine or mesorrhine. The southern Mediterranean belt is typified by moderately leptorrhine peoples; and in the eastern extremity, where aquilinity is the rule, extreme leptorrhiny is very common. The most leptorrhine area in Europe itself is the Dinaric region, particularly Montenegro and northern Albania, where mean nasal indices below 60 are encountered. In most of western Europe the noses are leptorrhine, but when one moves into Russia and the northeastern Balkans, mesorrhiny becomes the predominant form, and nasal indices increase perceptibly as one moves eastward, to a high mesorrhine or even platyrrhine level. Turkish-speaking peoples in the East, however, form an exception to this rule. Turkomans, Azerbaijans, and the like are, as a rule, extremely leptorrhine, more so than the inhabitants of Asia Minor and the Caucasus. On the opposite side of the map, the extreme western fringe of tall, large-headed, meso- to brachycephalic peoples is likewise characterized by a slight increase in the nasal index. The Palaeolithic survivors were not notably leptorrhine; they were, in fact, much less so than the Nordics and others who followed them.

If one were to study the form of the orbits and the shape of the external eye, with adequate data, a very interesting and significant distribution might be seen. For example, the distance between the eyes is relatively great among all of the Slavic and Finnish peoples of eastern Europe, and this dimension increases as one approaches mongoloid territory. It is of moderate size in almost all of northwestern and central Europe, but again becomes pronounced in Ireland, along the coast of Norway, and in the Alpine regions, where one may attribute this wide-eyed condition not to mongoloid influences but again to a Palaeolithic survival.

There are two zones of narrow inter-orbital diameters: (1) the entire Mediterranean zone from the Atlantic to India, and (2) the Dinaric zone reaching from the north of Italy to northern Greece. Again in the so-called Armenoid region of Anatolia and in Armenian territory itself. an extremely narrow inter-orbital distance prevails. This criterion may perhaps survive as a means of discrimination between facially characteristic Palaeolithic survivors and mongoloids, on the one hand, and basic Mediterraneans and Armenoid-Dinarics, on the other.

The size, robusticity, and general form of the lower law is again an excellent racial criterion, but there is not enough data to permit it to be plotted. The Mediterranean zone from Morocco to India is characterized by a light, shallow jaw, a narrow bigonial diameter, and a restricted height dimension between the lower dental border and chin. This is the typical Mediterranean mandible, whether one finds it in Spain or in Arabia. The heaviest jaws and greatest bigonial diameters are found in the northwestern European borderlands, and in eastern Europe, where mongoloid influence is strong. The relatively light, narrow jaw of many Dinarics and Armenoids again suggests that these types are for the most part brachycephalized forms of tall Mediterraneans.

Notes:

17 Page 243.

18 Du Bois, E., CRIC, 1934, pp. 71-75; also, Marett, J., p. 129.

19 Duckworth, W. L. H., JAPL, vol. 51, 1917, pp. 167-179.
Fischer, E., MAGW, vol. 36, 1906, pp. 54-57.
Gladstone, R. J., Biometrika, vol. 4, 1905/6, pp. 105-123.
Mies, J., MAGW, vol. 20, 1890, pp. 37-49.
Weisbach, A., MAGW, vol. 19, 1889, pp. 198-200.

20 Sailer, K., Die Fehmarner, DRK, vol. 4, 1930.

21 Unpublished data in author's possession.