1

Consider the two theorems (3.15 and 3.16) of Rudin's functional analysis:

Theorem 3.15: If $V$ is a neighborhood of $0$ in a topological vector space $X$ and if $$K = \left\{\Lambda \in X^* : |\Lambda x| \leq 1 \; \text{for every} \; x \in V \right\}$$ then $K$ is weak*-compact

Theorem 3.16: If $X$ is a separable topological vector space, if $K \subset X^*$, and if $K$ is weak*-compact then $K$ is metrizable, in the weak*-topology.

How are such theorems used to prove the following

If $V$ is a neighborhood of $0$ in a separable topological vector space $X$, and if $\left\{\Lambda_n \right\}$ is a sequence in $X^*$ such that $$|\Lambda_n x | \leq 1 \;\;\; (x \in V, n = 1,2,3,\ldots)$$ then there's a subsequence $\left\{ \Lambda_{n_i} \right\}$ and there's a $\Lambda \in X^*$ such that $$ \Lambda x = \lim_{i \to \infty} \Lambda_{n_i} x \;\;\; (x \in X) $$ In other words, the polar of $V$ is sequentially compact in the weak*-topology

I understand that $\Lambda_n$ belong to $K$, as defined in the theorem 3.15, also by theorem 3.16 such set is metrizable, but how do I get the conclusion?

user8469759
  • 5,285

1 Answers1

1

Since $K$ is metrizable and $w^*$-compact, it is sequentially $w^*$-compact. This means that for any sequence $\{\Lambda_n\}_n \subset K$ we can extract a subsequence such that $$ \Lambda_{n_k} \to \Lambda \quad \text{in weak$^*$ sense} $$ to some $\Lambda\in K$. Now, convergence in the weak$^*$ sense simply means that $$ \Lambda x =\lim_{k\to\infty}\Lambda_{n_k}x \quad \text{for all $x\in X$}. $$

BigbearZzz
  • 15,084
  • "Since $K$ is metrizable and w∗-compact, it is sequentially $w$∗-compact", why is that? – user8469759 Jan 16 '19 at 10:43
  • @user8469759 In a metric space topology compactness and sequentially compactness coincide. – BigbearZzz Jan 16 '19 at 10:44
  • The proof is not obvious but quite standard in most metric space textbook. For example, you can read about it here https://math.stackexchange.com/questions/822236/compactness-and-sequential-compactness-in-metric-spaces?noredirect=1&lq=1 – BigbearZzz Jan 16 '19 at 10:45
  • I was just reading through that link, thank you. I was wondering if "Munkres - topology" provides such proof. – user8469759 Jan 16 '19 at 10:46
  • @user8469759 Usually it's a part of the equivalent trilogy "compactness", "seq-compactness" and "complete&totally boundedness". I'm sure there's a proof of this in Intro Topology & Modern Analysis by Simmons. – BigbearZzz Jan 16 '19 at 10:48