Abstract In Italy there are two species of Apodemus (Sylvaemus): Apodemus sylvaticus on the mainland and the main island, and Apodemus flavicollis only on the mainland. The trend of some morphometric characters of the skull (incisive foramen length - FI; interorbital breadth = IO; length of palatal bridge = PP; upper alveolar length = $M^1M^3$ ) was analized and some theoretical models verified for A. sylvaticus. If one considers the sympatric population of A. sylvaticus and A. flavicollis simultaneously the characters PP, IO and $M^1M^3$ appear significantly isometric being directly correlated ($P \leq O.O1$), while FI character results allometric with respect to the previous ones, as expected. If one considers the sympatric populations of each of the species separately, the scenario is different. For A. sylvaticus only PP and $M^1M^3$ are isometric ($P \leq 0.05$). For A. flavicollis only $M^1M^3$ and FI appear to be correlated, although not as significantly as for A. sylvaticus ($P \le 0.05$; one tail). The insular populations of A. sylvaticus do not show significant correlations, except for FI and $M^1M^3$ ($P \le 0.05$). On the contrary, considering all populations, sympatric and allopatric, of A. sylvaticus at the same time are significant correlations ($P \le 0.05$) in all combinations of characters, except for those involving the IO. We suggest that the isometric relations in sympatric assemblages are confined within a morphological range available to the genus Apodemus. In such a space, the two species are split in two different and innerly homogeneous distributions. We found no evidence to confirm the niche variation hypothesis. On the contrary, the variability expressed as SO or CV's appears higher in the sympatric populations than in the allopatric ones, for three of the four characters, confirming previous results. For the character displacement hypothesis, we found that A. sylvaticus when not found together with A. flavicollis, shows adimensional increase of PP up to an intermediate value between the mean values of A. sylvaticus and A. flavicollis sympatric assemblages. Also $M^1M^3$, FI and FI + PP increase in allopatry, but to a value greater than that of A. flavicollis; on the other hand, IO growth in allopatry is not significant, as expected, because such character is allometric in the species under study. Lastly, the model was completely inadequate for FI, a character that in sympatry is higher in A. sylvaticus, but increases to a greater extent in the allopatric populations of A. sylvaticus. As far as the convergence on a given phenotype is concerned, the scarse available data are not in agreement with the model. The comparison between the sympatric A. sylvaticus and A. flavicollis shows a correlation of the IO and FI + PP value, when considered in the same localities ($P \le 0.05$). Similar observations were made by Tvrtkovic' (1979) on FI and $I-M^3$ (upper tooth length) for several european localities ("character's paralleling"). Moreover, for some isometric and allometric characters, the variability is higher when comparing the interspecific averages between the local and total values. This suggests that homotopic average of the cumulative dimension of the two species, is more variable than the interspecific homotopic differences. Possibly, A. sylvaticus, on the islands, where his congener is not present, tends to invade the morphological range of A. flavicollis. At A. sylvaticus specific level, all characters linked to the skull length appear isometric, whereas IO linked to skull breadth is allometric. Conversely, concerning heterospecific sympatric assemblages, the all pairs of characters linked to skull dimension are isometric, while FI (negatively linked to bones thickness and positively linked to skull length) is allometric. The intraspecific comparison perhaps reflects a gradient of skull length, whereas the interspecific one a difference in size and thickness between A. sylvaticus and A. flavicollis skull. On the other hand it appears, from Filippucci et al. (1984) that the importance of genetical factors is highly dependent on the particular character examined and on the species under study. In such a manner, results which are ecologically rather similar could be obtained with different processes. Conversely, the isometry and allopatry of characters seems to depend on different ways from the taxonomical, evolutive and ecologica1 context taken into account. Namely, some characters seem to be more subject to the competitive pressure, others to the genera1 structural model of the species. On the other hand, it is not sure that competition can explain all the observed phenomena, some of which may have a multiple base. More research is needed in order to elucidate such casual aspects of observed phenomena. Riassunto Su Apodemus sylvaticus dell'Italia insulare e su A. sylvaticus ed Apodemus flavicollis dell'Italia peninsulare sono stati studiati alcuni caratteri cranici, evidenziandone l'isometria o l'allometria. Su detti caratteri sono stati verificati e discussi i modelli eco1ogici: variazione di nicchia, spostamento dei caratteri, convergenza fenotipica e parallelismo di caratteri in condizione di simpatria e/o allopatria.