Axiom (Ov) is equivalent to the following:
(Ov′) If A and B are any two disjoint closed subsets of X, then there exist two disjoint open sets U,V such that A⊂U and B⊂V.
It follows immediately that (Ov) implies (Ov′), for if f is a continuous mapping of X into [0,1] which is equal to 0 on A and to 1 on B, then the open sets f−1([0,1/2[) and f−1(]1/2,1]) contain A and B respectively and do not intersect.
To prove the converse, notice first that (Ov′) is equivalent to the following axiom:
(Ov′′) Given any closed set A and any open neighbourhood V of A, there exists an open neighbourhood W of A such that W⊂V.
If there is a continuous mapping f:X→[−1,+1] which is equal to −1 on A and to +1 on B, and if we put U(t)=f−1([−1,t[) for each t∈[0,1], then we have defined a family of open sets in X, indexed by [0,1], such that (i) A⊂U(0), (ii) B⊂CU(1) and (iii) for each pair of real numbers t,t′ such that 0≤t<t′≤1 we have
U(t)⊂U(t′);
for U(t) is contained in the closed set f−1([−1,t]). Conversely, suppose that we have defined a family (U(t)) of open sets (0≤t≤1) with these three properties (i), (ii) and (iii). For each x∈X, put g(x)=1 if x∈CU(1), and if x∈U(1) let g(x) be the greatest lower bound of the values of t such that x∈U(t). Clearly 0≤g(x)≤1 for each x∈X, g(x)=0 on A, g(x)=1 on B; also g is continuous on X, for if we put g(x)=a, we have ∣g(y)−g(x)∣≤ε for all y∈U(a+ε)∩CU(a−ε), which is a neighbourhood of x by (i) [with the conventions that U(a+ε)=X if a+ε>1, and U(a−ε)=∅ if a−ε<0].
Hence Theorem 1 will be proved if we can define a family (U(t)) of open sets satisfying conditions (i), (ii) and (iii) above; to do this we use Axiom (OV′). Take U(1)=CB; since A⊂U(1) there exists an open set U(o) such that A⊂U(o) and U(o)⊂U(1) by (OV′). Suppose then that for each dyadic number k/2n (k=0,1,…,2n) we have defined an open set U(k/2n), these sets being such that U(k/2n)⊂U((k+1)/2n) for 0≤k≤2n−1. For each dyadic number
(2k+1)/2n+1(0≤k≤2n−1) there exists by
(OV′) an open set
U((2k+1)/2n+1) such that
U(k/2n)⊂U((2k+1)/2n+1) and
U((2k+1)/2n+1)⊂U((k+1)/2n). Hence for each dyadic number
r such that
0≤r≤1 we can define an open set
U(r), such that
A⊂U(o),
B⊂CU(1), and
U(r)⊂U(r′) for each pair of dyadic numbers
r,r′ such that
0≤r<r′≤1.
Now define, for each real number t∈[0,1],
U(t)=r≤t⋃U(r)(r dyadic); by (2), this definition agrees with the preceding one for
t dyadic; also, if
0≤t<t′≤1, then there exist two dyadic numbers
r,r′ such that
t≤r<r′≤t′; by (2) we have
U(t)⊂U(r), hence
U(t)⊂U(r)⊂U(r′)⊂U(t′); this proves (1) and therefore completes the proof.
Theorem 1 will enable us to show that two important categories of topological spaces are normal. In the first place:
Bourbaki puts the proof directly after the statement, so what is above is the statement and its proof together.
Nothing.