1: \begin{abstract}

2: We prove that if $X$ is a product of elliptic curves over a

3: finite field $k$, rational and numerical equivalences agree on $X$. The

4: proof works for any smooth projective variety $X$ ``of abelian type" for

5: which the Tate conjecture holds (the case of products of elliptic curves

6: being due to M. Spie\ss's \protect\cite{spiess}). It uses Soul\'e's ideas

7: \protect\cite{soule}, U. Jannsen's semi-simplicity theorem

8: \protect\cite{jannsen},   and a result of Y. Andr\'e and the

9: author \protect\cite{ak} inspired by S.I. Kimura's results on

10: finite-dimensional Chow motives \protect\cite{ki}. We give some

11: consequences, among which: the conjectures of Lichtenbaum

12: \protect\cite[\S 7]{licht} hold true for $X$,

13: the second Chow group of $X$ is finitely generated, the Beilinson-Soul\'e

14: conjecture holds in weight $n$ for the function field of $X$ provided

15: $n\le 2$ or $\dim X\le 2$, Gersten's conjecture holds for discrete

16: valuation rings with residue field such a function field if $\dim X\le 2$.

17: Also, if

18: $U$ is an open subset of

19: $X$ with

20: $\dim X\le 2$, the action of Frobenius on $H^*_c(\bar U,\Q_l)$ is

21: semi-simple and its characteristic polynomial does not depend on $l\ne

22: \text{ char } k$.

23: \end{abstract}

24: