1: \begin{abstract}

2:   This brief pedagogical note re-proves a simple theorem on the convergence, in

3:   $L_2$ and in probability, of time averages of non-stationary time series to

4:   the mean of expectation values.  The basic condition is that the sum of

5:   covariances grows sub-quadratically with the length of the time series.  I

6:   make no claim to originality; the result is widely, but unevenly, spread bit

7:   of folklore among users of applied probability.  The goal of this note

8:   is merely to even out that distribution.

9: \end{abstract}

10: