[guided]We need to show $\Gamma: A \to C(\Phi_A)$ is **surjective**. The strategy is to identify $\hat{A} = \Gamma(A)$ as a subset of $C(\Phi_A)$ and apply the Stone-Weierstrass theorem to conclude $\hat{A}$ is **dense**, and then use isometry of $\Gamma$ to conclude $\hat{A}$ is **closed** — together giving $\hat{A} = C(\Phi_A)$.
**The Stone-Weierstrass theorem (complex version).** Let $K$ be compact Hausdorff. A subset $\mathcal{B} \subseteq C(K)$ is dense in $(C(K), \|\cdot\|_\infty)$ provided:
1. $\mathcal{B}$ is a $*$-subalgebra (closed under $+$, scalar $\cdot$, pointwise product, and complex conjugation),
2. $\mathcal{B}$ is unital (contains the constant function $1$),
3. $\mathcal{B}$ separates points of $K$.
We verify these for $\mathcal{B} = \hat{A}$ and $K = \Phi_A$.
**Why $\hat{A}$ is a subalgebra.** $\Gamma$ is a unital algebra homomorphism (this is general Gelfand theory, theorem 2678): $\Gamma(x + y) = \Gamma(x) + \Gamma(y)$, $\Gamma(xy) = \Gamma(x)\Gamma(y)$, $\Gamma(1_A) = 1_{C(\Phi_A)}$. So the image is a unital subalgebra.
**Why $\hat{A}$ is conjugation-closed.** Here is where Step 2 enters: $\widehat{x^*} = \overline{\hat{x}}$. This says: the complex conjugate of any element of $\hat{A}$ is again in $\hat{A}$ (it's the Gelfand transform of $x^*$). *Without the previous theorem* (**Characters Are Star-Homomorphisms**), this property would not be available; the Gelfand theory of *general* commutative Banach algebras gives a homomorphism $\Gamma$, but not a $*$-homomorphism.
**Why $\hat{A}$ separates points of $\Phi_A$.** Two distinct characters $\varphi, \psi$ are *by definition* distinct as linear functionals on $A$, so there exists $x \in A$ with $\varphi(x) \neq \psi(x)$. The function $\hat{x}: \Phi_A \to \mathbb{C}$, $\varphi \mapsto \varphi(x)$, separates them: $\hat{x}(\varphi) \neq \hat{x}(\psi)$. This is **automatic** for the Gelfand topology — the weak-$*$ topology on $\Phi_A$ is exactly the topology making all the $\hat{x}$ continuous, so distinct points are distinguished by some $\hat{x}$.
**Conclusion of Stone-Weierstrass.** $\hat{A}$ is dense in $C(\Phi_A)$.
**Why $\hat{A}$ is closed.** Step 3 gives $\|\Gamma(x)\|_\infty = \|x\|$ for all $x \in A$. So $\Gamma: A \to \hat{A}$ is an isometric bijection. Since $A$ is complete (Banach), the image $\hat{A}$ is complete, hence closed in $C(\Phi_A)$ (a complete subset of a metric space is closed).
**Density + closedness = surjectivity.** A dense closed subset of any topological space is the whole space. Hence $\hat{A} = C(\Phi_A)$.
**Where the $C^*$-axiom enters.** The whole argument breaks for general commutative Banach algebras. The two crucial $C^*$-specific inputs are:
- *$*$-preservation* (Step 2), needed to show $\hat{A}$ is conjugation-closed. Stone-Weierstrass requires conjugation-closure to handle complex-valued functions; without it we would need to work over $\mathbb{R}$ and miss the imaginary parts.
- *Isometry* (Step 3), needed to show $\hat{A}$ is closed. In a general Banach algebra, $\|\hat{x}\|_\infty = r_A(x)$ can be strictly less than $\|x\|$ (e.g.\ a non-zero nilpotent has $r_A = 0$ but positive norm), so $\Gamma$ would be strictly contractive and the image would not be closed.
It is precisely the convergence of these two $C^*$-features — both flowing from the $C^*$-identity $\|x^*x\| = \|x\|^2$ — that makes $\Gamma$ an isometric $*$-isomorphism, and hence makes $A$ literally the algebra of continuous functions on its character space.[/guided]