1: \begin{abstract}

2: Chaos is widely

3: understood as being a consequence of sensitive dependence 

4: upon initial conditions. This is the result of an instability

5: in phase space, which separates trajectories exponentially.

6: Here, we demonstrate that this criterion should be refined.

7: Despite their overall intrinsic instability, trajectories may be very strongly

8: convergent in phase space over extremely long periods, as revealed by

9: our investigation of a simple chaotic system (a realistic model for

10: small bodies in a turbulent flow).

11: We establish that this strong convergence is a

12: multi-facetted phenomenon, in which the clustering is intense, widespread and

13: balanced by lacunarity of other regions.

14: Power laws, indicative of scale-free features, characterize

15: the distribution of particles in the system. 

16: We use large-deviation and extreme-value statistics to explain the effect. 

17: Our results show that the interpretation of the \lq butterfly 

18: effect' needs to be carefully qualified. We argue that the 

19: combination of mixing and clustering processes makes our specific model 

20: relevant to understanding the evolution of simple organisms. Lastly, this 

21: notion of \emph{convergent chaos}, which implies the existence of conditions 

22: for which uncertainties are unexpectedly small, may also be relevant

23: to the valuation of insurance and futures contracts.  

24: \end{abstract}

25: