a) For every proper partial mapping φ of X into Y, the mapping f↦f∘φ of C(Y′) into C(X′) induces an algebra morphism φ∗ of C0(Y) into C0(X) ;
b) The mapping φ↦φ∗ is a bijection of the set of proper partial mappings of X into Y onto the set of algebra morphisms of C0(Y) into C0(X). Its inverse bijection is the mapping π↦X′(π).
Let us prove a). Let φ be a proper partial mapping of X into Y, identified with a continuous mapping of X′ into Y′ such that φ(ωX)=ωY. For f∈C0(Y), one has (f∘φ)(ωX)=f(ωY)=0, hence the mapping φ∗ is well-defined. It is an algebra morphism.
Let us prove that X′(φ∗) identifies with φ. Let x∈X. For every function f∈C0(Y), the character X′(φ∗)(evx) associates to f the complex number
(evx∘φ∗)(f)= evx(f∘φ)=f(φ(x)),
hence X′(φ∗)(evx) = evφ(x). This proves the assertion.
Conversely, let π:C0(Y)→C0(X) be an algebra morphism. Let us prove that X′(π)∗=π. Let f∈C0(Y), and let us write g=X′(π)∗(f)∈ C0(X). For every x∈X, one has
g(x)=(f∘X′(π))(x)= evX′(π)(x)(f)= (evx∘π )(f)=π(f)(x),
since X′(π) satisfies evX′(π)(x)= evx∘π. Hence g=π(f), which permits us to conclude that X′(π)∗=π.
In an entirely similar manner, one has:
Bourbaki sets the proof directly after the statement, and the corpus does not separate them, so what is above is the statement and its proof together.
Nothing.
Nothing.