Zalai Múzeum 8. (Zalaegerszeg, 1997)
Szathmáry László: Divergency of the Celts and Scythians in the Territory of Hungary
100 Szathmáry László Method Twelve of the cranial measurements were used for the examinations: Greatest cranial length (after Martin - 1928 -, No. 1=M1) Nasion-Basion length (M5) Greatest breadth (M8) Minimum frontal breadth (M9) Basion-Bregma height (Ml 7) Porion-Bregma height (M20) Bizygomatic breadth (M45) Upper facial height (M48) Orbital width (M51) Orbital height (M52) Nasal width (M54) Nasal height (M55) Pet. of correctly classified cases: 93,3%. It is very surprising that the Scythian and the Celtic skulls present extremely different variations. In the case of females, nearly each of the 10 Celtic cranial finds manifest, a specific habitus while only 8 out of the 45 Scythian skulls show some similarity to the Celtic sample. Consequently, the number of the correctly classified cases is really great: 83,6 per cent. With the males this percentage is somewhat higher: 93,4. 17 out of the 19 Celtic Endings are typical of their own sample. And, at the same time, 39 out of the 41 Scythian skulls are typical of their own sample. This Because of the too many missing data in the case of females, the measurements M20 and M48 had to be omitted and this way only 10 dimensions could be made use of. The estimation of missing data was performed by Dear's method (1959). The two samples were compared according to sexes by applying a discriminant analysis. Results The diversity of the sample composed of the two groups can be characterised by one significant canonic discriminant function. The two paleoethnic components can be most clearly separated according to the two orbital dimensions (M5 1 and M52) as well as to the nasion-basion lenght (M5). clear manifestation of harmony is definitely unusual with samples of such a small number. That is why we may assume with good reason that both ethnic groups may have had their own special craniological profile. In this way, if an Iron Age finding without grave-goods got dug up, we have a fair chance of deciding successfully which of the two above mentioned groups it may be assigned to. Now, it is the classification results that are the most interesting for us: Classification results — Females Actual group Cases Predicted groups Celts Scythians Celts Scythians 10 45 9 (90,0%) 8 (17,8%) 1 (10,0%) 37 (82,2%) Pet. of correctly classified cases: 83,6%. Classification results — Males Actual group Cases Predicted groups Celts Scythians Celts Scythians 19 41 17 (89,5%) 2 (4,9%) 2 (10,5%) 39 (95,1%)