Six independent traits, some old, some new, were examined, occipital index (Fig. 2), sphenomaxillary suture (Fig. 3b), transverse zygomatic suture trace (Fig. 4), infraorbital pattern type II (Fig. 5), transverse palatine suture index (Fig. 6), and supraorbital foramen (Fig. 7). These characters may presently be regarded as optimal for racial taxonomy since they remained after many others were discarded as useless. Their frequencies in 67 recent and 12 prehistoric cranial series (Tab. 1, 2; Fig. 1) were subjected to principal components analysis at various levels.

At the world level (Tab. 3, 4; Fig. 8), ist 1st principal component (PC1) sets Europoids apart from Mongoloids while PC2 separates the Jomon population from the rest of mankind and is very useful for evaluating the proportion of aboriginal and immigrant genes in the populations of Japan. Hokkaido Ainu appear to be direct descendants of the Jomonese. No similarity is seen between Africans and Australoids. Amerindians and East Asians lack the Mongoloid specialization peculiar to North Asians.

When only Europoids, Mongoloids, and groups traditionally regarded as hybrid, are included in the analysis (Tab. 4, 5; Fig. 9), PC1 reveals a clear-cut west to east cline in Eurasia (Fig. 10), and PC2 differentiates several Uralian speaking peoples (Lapps, Volga Finns, Ob Ugrians, Samoyeds) and Yukagirs from the remaining populations (Fig. 11) thus supporting the idea of a Proto-Uralian substratum. Eskoaleuts and Chukchi bridge the gap between Amerindians and Siberians.

Within the Western sample (Tab. 4, 6; Fig. 12), Northern Europoids differ from Southern Europoids. Within the Eastern sample, (Tab. 4, Fig. 13), clusters based on epigenetic data are in excellent agreement with the linguistical classification.