1: \begin{abstract}

2: We discuss a notion of convergence for binary trees that is based on

3: subtree sizes. In analogy to recent developments in the theory of graphs, posets and

4: permutations we investigate some general 

5: aspects of the topology, such as a characterization of the set of possible 

6: limits and its structure as a metric space. For random trees the subtree 

7: size topology arises in the context of algorithms for searching and sorting 

8: when applied to random input, resulting in a sequence of nested trees. For these

9: we obtain a structural result based on a local version of exchangeability. 

10: This in turn leads to a central limit theorem, with possibly mixed asymptotic 

11: normality.

12: \end{abstract}