Bourbaki

Proposition 2

Tag 0BOK, permanent. Cite it as [Bourbaki, Tag 0BOK].

General Topology, chapter III, § 7. Inverse limits of topological groups and rings

Let GG be a topological group and let (Hα)α∈I(H_\alpha)_{\alpha \in I} be a family of normal subgroups of GG such that Hα⊃HβH_\alpha \supset H_\beta whenever α≤β\alpha \leq \beta, and which satisfy the following condition:

(AP) For each α∈I\alpha \in I, HαH_\alpha is closed in GG and every neighbourhood of the identity element ee in GG contains one of the HαH_\alpha (in other words, the filter base formed by the HαH_\alpha converges to ee).

Then the mapping i:G→G~=lim←⁡G/Hαi : G \to \tilde{G} = \varprojlim G / H_\alpha is a strict morphism of GG onto i(G)i(G); G~\tilde{G} is Hausdorff and i(G)i(G) is dense in G~\tilde{G}; and finally the kernel of ii is the closure of {e}\{e\} in GG. If in addition one of the HαH_\alpha is complete, then ii is surjective.

Clearly the Gα=G/HαG_\alpha = G / H_\alpha are Hausdorff (§ 2, no. 5, Proposition 13), hence so is G~\tilde{G} (since it is a subspace of ∏α∈IGα\prod_{\alpha \in I} G_\alpha). The kernel HH of ii is the intersection of the HαH_\alpha and is therefore a closed subgroup of GG. Since each neighbourhood of ee contains some HαH_\alpha, it contains HH and therefore [§ 3, no. 1, formula (1)] HH is the closure of {e}\{e\}. Let us next show that i(G)i(G) is dense in G~\tilde{G}. Let fαf_\alpha be the canonical mapping G~→Gα\tilde{G} \to G_\alpha, which is the restriction to G~\tilde{G} of the projection prαpr_\alpha; φα=fα∘i\varphi_\alpha = f_\alpha \circ i is the canonical mapping G→G/HαG \to G/H_\alpha. If UU is any non-empty open set of G~\tilde{G}, then there is an index α∈I\alpha \in I and a non-empty open set UαU_\alpha in GαG_\alpha such that f‾α−1(Uα)⊂U\overline{f}_\alpha^{-1}(U_\alpha) \subset U (Chapter I, § 4, no. 4, Proposition 9); therefore

i‾−1(U)⊃φ‾α−1(Uα); \overline{i}^{-1}(U) \supset \overline{\varphi}_\alpha^{-1}(U_\alpha);
but since φα\varphi_\alpha is surjective, i‾−1(U)\overline{i}^{-1}(U) is not empty, and therefore i(G)∩U≠∅i(G) \cap U \neq \emptyset.

To see that ii is a strict morphism of GG onto i(G)i(G), consider a neighbourhood VV of ee in GG. There is a neighbourhood WW of ee in GG such that W2⊂VW^2 \subset V, and an index α∈I\alpha \in I such that Hα⊂WH_\alpha \subset W; it follows that VV contains WHα=φ‾α(φα(W))=i‾−1(f‾α−1(φα(W)))WH_\alpha = \overline{\varphi}_\alpha(\varphi_\alpha(W)) = \overline{i}^{-1}(\overline{f}_\alpha^{-1}(\varphi_\alpha(W))). Since f‾α(φα(W))\overline{f}_\alpha(\varphi_\alpha(W)) is a neighbourhood of the identity element in G~\tilde{G}, the result follows.

Finally, suppose that there is an index γ∈I\gamma \in I such that HγH_\gamma is complete. To show that ii is surjective it is enough to prove that every family (Tα)α∈I∈G~(T_\alpha)_{\alpha \in I} \in \tilde{G} has a non-empty intersection. Since TγT_\gamma is obtained by translation from HγH_\gamma, it is a complete subspace of GG (with respect to both the right and the left uniformities). Moreover, since every neighbourhood UU of ee in GG contains one of the HαH_\alpha, the corresponding set TαT_\alpha is Ud−U_{d-} (or Us−U_{s-}) small, and hence the set of TαT_\alpha contained in TγT_\gamma is a Cauchy filter base; therefore it converges in TγT_\gamma, and since the TαT_\alpha are closed in GG (since they are translations of the HαH_\alpha), their intersection is not empty.

Q.E.D.

Bourbaki puts the proof directly after the statement, so what is above is the statement and its proof together.

Cited by

Cites

Transcribed from the printed volume. Not translated.