1: \begin{abstract}

2: We consider the Voronoi tessellation based on a homogeneous Poisson point 

3: process in $\mathbf{R}^{d}$. For a geometric characteristic of the cells

4: (e.g. the inradius, the circumradius, the volume), we investigate the point

5: process of the nuclei of the cells with large values. Conditions are

6: obtained for the convergence in distribution of this point process of

7: exceedances to a homogeneous compound Poisson point process. We provide a  

8: characterization of the asymptotic cluster size distribution which is based

9: on the Palm version of the point process of exceedances. This

10: characterization allows us to compute efficiently the values of the extremal index and the cluster size probabilities by simulation for various geometric characteristics. The extension to the Poisson-Delaunay

11: tessellation is also discussed.

12: \end{abstract}

13: