# A question about infinite utility streams

At the end of Diamond's Evaluation of Infinite Utility Streams he proves a theorem (which he doesn't give a name to, but it's at the very end of the article). There is a step in which he jumps from $(u,0)_{rep}\succ (0,u)_{rep}$ to $(u,0)\succ_t (0,u)$, and I don't understand where that comes from. It seems like it's the opposite direction of axiom A2, and I don't see how that's derived.

-

Step: $(u,0)_{rep}\succ (0,u)_{rep}\implies (u,0)\succ_2(0,u)$
Proof: Suppose not. Since $\succeq_2$ is complete, we would have $(0,u)\succeq_2 (u,0)$ otherwise, and hence by A2 $(0,u)_{rep}\succeq (u,0)_{rep}$. This cannot be.
I don't see how the rest of Diamond's proof works out though. $(u,0,U)\succeq(0,u,U)$ follows now from A1, but I don't see why this should be strict.