Vörös A. szerk.: Fragmenta Mineralogica Et Palaentologica 16. 1993. (Budapest, 1993)

pattern of Miocene cricetids (Daxner-Höck 1972). The presence of rudimentary mesolophids is ancestral feature too (Fig. 10, c). On the scatter diagram of the ml molars a part of the teeth from Esztramos 3 is smaller sized but not separated from the mass of the other molars. m 2 Morphological distribution: A: ALC+, ML+ 12 B: ALC-, ML+ 16 C: ALC+, ML19 D: ALC-, ML7 In the dental pattern the presence of mesolophid and the antero-lingual cingulum (ALC) is much more frequent than in the Pleistocene materials (Fig. 11). On the scatter diagram (Fig. 5) a weakly separated large sized group is visible. In the distribution of the Wa-Wp there is not any remarkable difference from the cited Polish materials contrast with the Ml - ml molars. m 3 Morphological distribution: A: ALC+, ML+ 29 G: ALC+, ML1 X: ALCML+ 15 E: ALCML3 The ALC is frequent and well developed (Fig. 12). The antero-buccal cingulum is constant. There is only one molar in the Esztramos 3/8 sample with week ABC (Fig. 12, d). The building up of the mesolophid region is very variable (Fig. 12). CONCLUSIONS The Allocricetus ehiki populations of Villány 3 and Esztramos 3 have some ancestral features in the morphology (mesolophe on Ml, mesolophid on ml, undivided or 3-parted anteroconid, frequent antero-lingual cingula on m2-3) differentiate it from Pleistocene Allocricetus. The most important metrical features are the next. The Ml and ml teeth are more trapezoidal than the same molars in the Polish Pleistocene and Pliocene and in the recent Cricetulus material. The assemblage of Villány 3 is rather homogeneous but the size distribution of teeth from Esztramos 3 is wider or bimodal (e.g. m2, Fig. 5). The two-grouped distribution is visible on the diagram of the Lml-Lm3 relation too (Fig. 7). As it appears to me the tendency of the separation of Allocricetus bursae from Allocricetus ehiki begins in the fauna of Esztramos 3. On this ground the slightly younger age of Esztramos 3 is possible. It was supposed also by Jánossy (1979,1986) on the basis of the faunistical composition of the two localities.