Abstract: In an article recently published, one of the authors (Wang Linghong et al. 1983) introduced a new fossil hominid find from China dated probably to the late Upper Pleistocene. This find, which was discovered in Huanglong County in 1975, consists of a part of the frontal and adjoining sections of the parietals. Although the vault fragment might belong to anatomically modern Homo sapiens, there are also some characteristics which could probably point to a more archaic type of anatomically modern Homo sapiens. In order to more extensively study the affinities of the Huanglong specimen, a multivariate statistical comparison was carried out in the present study which used recent and Upper Pleistocene cranial material.
In order to compare the Huanglong specimen with a rather broad morphological spectrum, numerous European and South-West Asian finds of the I'pper Pleistocene were included along with the Chinese material. Such an Asian-European comparison is of great interest not only for the analysis of the affinities of tlie Huanglong hominid, but also with regard to the affinities between Maba and the western Neandertals as well as between ''Zhoukoudian, Upper Cave'' and the Cro-Magnon type.
In the selection of the variables and the statistical method, the analysis of the Upper Pleistocene frontal from Hahnofersand (near Hamburg, West Germany) was useful (Brauer 1980, 1981). In this study the same ten variables (see Fig. 1 and Table 2) and the principal components method (Nie et al. 1975) were used.
Two calculations with different groups of individuals were carried out. All of the variables were included in both. In the first analysis, the Huanglong specimen was compared only to the other Chinese material. Table 2 presents the values of the variables of the Chinese specimens. Two modern Chinese specimens (Nos. 94 and 100) which differ in many respects were chosen for this study. In the second calculation, thirty representatives of European and West-Asian Neadertals and Upper Paleolithic Europeans were included along with the Chinese material (cf. Table 1). For both analyses, the first two principal components already explain a very large part of the total variance (analysis 1: 94.1% ； analysis 2: 88.5%)； thus only these two components have to be treated here.
Figure 3 shows the distribution of the six Chinese individuals with regard to the first two components. While the linear measurements of the frontal profile have generally high loadings on the first component, the variables "minimum frontal width" and the two angles are mainly represented by the second component. The most extreme positions on the first component are held by the two modern Chinese specimens, which thus also encompass the values of all other finds, including Maba.
Concerning the angles of the curvature of the frontal (component 2), Huanglon^ holds a somewhat more archaic position compared to Liujiang and thus shows certaia affinities to the frontal of Maba.
A more complex picture, which allows better interpretations, results when a larger number of finds from Europe and West Asia are included, as in analysis 2 (Fig. 4).
If one compares the Chinese specimens to this spectrum, then the large differences between the two modern Chinese become evident. While the frontal No. 94 shows very close relationships to the more gracile finds from Ofnet, No. 100 shows affinities to the robust specimens from Predmost and Engis.
The position which Huanglong occupies on the basis of the first two components is very interesting. This find deviates a little from the modern spectrum towards the archaic direction ； its position can thus not be regarded as intermediate between the Neandertals and anatomically modern Homo sapiens. There are stronger relationships to modern Homo sapiens, which are expressed in the specific affinities to Combe-Ca- pelle. Upper Cave No. 101 also deviates in the same direction from the Upper Paleolithic Europeans, to which it nevertheless possesses the greatest similarity (e.g. to Obe- rkassel cf). Liujiang exhibits no differences to the Upper Paleolithic Europeans in this analysis. Thus, this analysis of frontal morphology supports the similarities between the late Pleistocene finds from China and Europe, which have often been determined either evolutionarily or racially. The position of Maba in Fig. 4 is quite similar to that of Upper Cave No. 101. Although Maba holds a somewhat more intermediate position between the Neandertals and the Upper Paleolithic spectrum, its relationship to the modern frontal shape is greater than to the flat Neandertaloid form. This is also shown by the fact Qafzeh VI clearly lies closer to the Neandertal spectrum than does Maba. The archaic character of Maba is primarily determined by the prominent supraorbital torus rather than by the frontal curvature.
Summing up the results of the two analyses with regard to the affinities of Huang- long, it is probable that the frontal morphology of this find might be somewhat more archaic than that of the modern Chinese. On the basis of this analysis which included only a few modern Chinese specimens, the possibility cannot be excluded that Huanglong still falls within the limits of modern variation or at least lies very close to them. Moreover, the comparison with European finds has shown that Huanglong appears to have stronger affinities to various Upper Paleolithic Europeans than to the two recent Chinese. The other specimens from the Upper Cave, however, as well as those from Liujiang and Ziyang, exhibit descriptive details on the skulls which have been regarded Proto-Mongoloid (Woo 1959, Wu 1961, Wolpoff et a]. 1984). Although this appears quite probable, one can not disregard the strong affinities which the total morphological pattern of these crania exhibits to the late Upper Pleistocene spectrum from other parts of the world.

Key words: Huanglong Calva; Anatomically modern Homo sapiens； archaic Homo sapiens； Principal components analysis