1: \begin{abstract}

2: In this paper, we consider multistopping problems for finite discrete

3: time sequences $X_1, \ldots , X_n$.

4: $m$-stops are allowed and the aim is to maximize the expected value of

5: the best of these $m$ stops.

6: The random variables are neither assumed to be independent not to be

7: identically distributed. The basic assumption

8: is convergence of a related imbedded point process to a continuous time

9: Poisson process in the plane,

10: which serves as a limiting model for the stopping problem. The optimal

11: $m$-stopping curves for this limiting model are

12: determined by differential equations of first order. A general

13: approximation result is established which ensures

14: convergence of the finite discrete time $m$-stopping problem to that in

15: the limit model. This allows the construction of approximative

16: solutions of the discrete time $m$-stopping problem. In detail, the

17: case of i.i.d. sequences with

18: discount and observation costs is discussed and explicit results are obtained.

19: \end{abstract}