[proofplan] We translate the vector-calculus identity into the language of differential forms and reduce it to the [Generalised Stokes' Theorem](/theorems/3555). To $F$ we associate the **work $1$-form** $\omega_F = F_1\, dx_1 + F_2\, dx_2 + F_3\, dx_3 \in \Omega^1(U)$. Two correspondences then carry the classical identity onto the form-theoretic one: (i) the circulation $\oint_{\partial S} F \cdot \tau \, d\mathcal{H}^1$ equals $\int_{\partial S} \omega_F$ by direct evaluation of $\omega_F$ on the positively-oriented unit tangent; (ii) the [exterior derivative](/theorems/1525) $d\omega_F$ corresponds to the curl via [Exterior Derivative Recovers Gradient, Curl, and Divergence on $\mathbb{R}^3$](/theorems/3566), and the resulting flux $2$-form integrates against the oriented surface to give $\int_S (\nabla \times F) \cdot n \, d\mathcal{H}^2$. Applying [Generalised Stokes' Theorem](/theorems/3555) to $\omega_F$ on $S$ finishes the proof. [/proofplan] [step:Associate the work $1$-form to the vector field $F$] Define the **work $1$-form** of $F$ as \begin{align*} \omega_F : U &\to \Lambda^1 T^*\mathbb{R}^3 \\ x &\mapsto F_1(x)\, dx_1\big|_x + F_2(x)\, dx_2\big|_x + F_3(x)\, dx_3\big|_x. \end{align*} Since each $F_i \in C^\infty(U)$, we have $\omega_F \in \Omega^1(U)$. At each $p \in U$ and each $v = (v_1, v_2, v_3) \in T_p\mathbb{R}^3 \cong \mathbb{R}^3$, \begin{align*} (\omega_F)_p(v) \;=\; F_1(p) v_1 + F_2(p) v_2 + F_3(p) v_3 \;=\; F(p) \cdot v. \end{align*} [guided] The first move in turning a vector-calculus identity into a form-theoretic one is to package the vector field $F$ as a covariant object that can be integrated over a curve. The natural choice is the $1$-form $\omega_F$ obtained by lowering the index using the standard Euclidean metric on $\mathbb{R}^3$. Concretely, we **declare** \begin{align*} \omega_F : U &\to \Lambda^1 T^*\mathbb{R}^3 \\ x &\mapsto F_1(x)\, dx_1\big|_x + F_2(x)\, dx_2\big|_x + F_3(x)\, dx_3\big|_x, \end{align*} and since the component functions $F_i$ are smooth on $U$, this is an element of $\Omega^1(U)$. Why is this the "right" correspondence? Because pairing $\omega_F$ with a tangent vector reproduces the Euclidean inner product. At any $p \in U$ and any $v = (v_1, v_2, v_3) \in T_p\mathbb{R}^3 \cong \mathbb{R}^3$, applying $\omega_F$ at $p$ to $v$ gives \begin{align*} (\omega_F)_p(v) \;=\; \sum_{i=1}^{3} F_i(p)\, dx_i|_p(v) \;=\; \sum_{i=1}^{3} F_i(p) v_i \;=\; F(p) \cdot v. \end{align*} This identity is the bridge: every "$F \cdot (\,\cdot\,)$" expression in the classical formulation will be rewritten as an evaluation of $\omega_F$. [/guided] [/step] [step:Identify the circulation $\oint_{\partial S} F \cdot \tau \, d\mathcal{H}^1$ with $\int_{\partial S} \omega_F$] The boundary $\partial S$ is a compact, oriented, smooth $1$-dimensional submanifold of $\mathbb{R}^3$, with continuous positively-oriented unit tangent field $\tau: \partial S \to \mathbb{R}^3$. By the definition of [Integration of Differential Forms](/theorems/1529) on an oriented $1$-manifold, \begin{align*} \int_{\partial S} \omega_F \;=\; \int_{\partial S} (\omega_F)_p(\tau(p)) \, d\mathcal{H}^1(p). \end{align*} Using the pointwise identity $(\omega_F)_p(\tau(p)) = F(p) \cdot \tau(p)$ from the previous step, \begin{align*} \int_{\partial S} \omega_F \;=\; \int_{\partial S} F(p) \cdot \tau(p) \, d\mathcal{H}^1(p) \;=\; \oint_{\partial S} F \cdot \tau \, d\mathcal{H}^1. \end{align*} [guided] We have already encoded the vector field $F$ as a $1$-form $\omega_F$. We now claim that the standard line integral $\oint_{\partial S} F \cdot dr$, written more carefully as $\oint_{\partial S} F \cdot \tau \, d\mathcal{H}^1$ where $\tau$ is the positively-oriented unit tangent to $\partial S$, is **literally** the integral of the $1$-form $\omega_F$ over the oriented curve $\partial S$. To see this, recall the definition of the integral of a $1$-form along an oriented compact $1$-submanifold: by [Integration of Differential Forms](/theorems/1529), for the oriented $1$-manifold $(\partial S, \tau)$ we have \begin{align*} \int_{\partial S} \omega_F \;=\; \int_{\partial S} (\omega_F)_p(\tau(p)) \, d\mathcal{H}^1(p). \end{align*} This is the **defining** formula: the $1$-form is evaluated on the unit positive tangent and the resulting scalar function is integrated against arc-length measure $\mathcal{H}^1$. Substituting the identity from the previous step, $(\omega_F)_p(\tau(p)) = F(p) \cdot \tau(p)$, we obtain \begin{align*} \int_{\partial S} \omega_F \;=\; \int_{\partial S} F(p) \cdot \tau(p) \, d\mathcal{H}^1(p) \;=\; \oint_{\partial S} F \cdot \tau \, d\mathcal{H}^1. \end{align*} The right-hand side is precisely the right-hand side of the Classical Stokes statement, written in the standard vector-calculus shorthand "$F \cdot dr$". So the line integral and the form integral coincide. [/guided] [/step] [step:Compute $d\omega_F$ and identify it with the curl $2$-form] By the [Coordinate Formula for the Exterior Derivative](/theorems/3564) applied to $\omega_F = \sum_{i=1}^{3} F_i\, dx_i$, \begin{align*} d\omega_F \;=\; \sum_{i=1}^{3} dF_i \wedge dx_i \;=\; \sum_{i,j=1}^{3} \partial_{x_j} F_i \, dx_j \wedge dx_i. \end{align*} Using $dx_j \wedge dx_i = -\, dx_i \wedge dx_j$ for $i \ne j$ and $dx_i \wedge dx_i = 0$, the sum collapses to \begin{align*} d\omega_F \;=\; (\partial_{x_2} F_3 - \partial_{x_3} F_2)\, dx_2 \wedge dx_3 \;+\; (\partial_{x_3} F_1 - \partial_{x_1} F_3)\, dx_3 \wedge dx_1 \;+\; (\partial_{x_1} F_2 - \partial_{x_2} F_1)\, dx_1 \wedge dx_2. \end{align*} Writing $G := \nabla \times F = (G_1, G_2, G_3)$ with $G_1 = \partial_{x_2} F_3 - \partial_{x_3} F_2$, $G_2 = \partial_{x_3} F_1 - \partial_{x_1} F_3$, $G_3 = \partial_{x_1} F_2 - \partial_{x_2} F_1$, this reads \begin{align*} d\omega_F \;=\; G_1\, dx_2 \wedge dx_3 + G_2\, dx_3 \wedge dx_1 + G_3\, dx_1 \wedge dx_2 \;=:\; \star_3 (\nabla \times F), \end{align*} where the right-hand side is the **flux $2$-form** associated to $\nabla \times F$. This is the content of [Exterior Derivative Recovers Gradient, Curl, and Divergence on $\mathbb{R}^3$](/theorems/3566). [guided] We need to identify the left-hand side of Classical Stokes — the flux integral of $\nabla \times F$ — with the integral of some $2$-form. Generalized Stokes says $\int_S d\omega_F = \int_{\partial S} \omega_F$, so the natural candidate is $d\omega_F$. We compute it explicitly. By the [Coordinate Formula for the Exterior Derivative](/theorems/3564), applied to a sum of products of smooth functions and coordinate $1$-forms, \begin{align*} d\omega_F \;=\; d\!\left(\sum_{i=1}^{3} F_i\, dx_i\right) \;=\; \sum_{i=1}^{3} dF_i \wedge dx_i \;=\; \sum_{i,j=1}^{3} \partial_{x_j} F_i \, dx_j \wedge dx_i. \end{align*} The wedge product is alternating: $dx_j \wedge dx_i = -\, dx_i \wedge dx_j$ when $i \ne j$, and $dx_i \wedge dx_i = 0$. The diagonal terms $i = j$ all vanish; the six off-diagonal terms pair up into three differences. Collecting them: \begin{align*} d\omega_F \;=\; (\partial_{x_2} F_3 - \partial_{x_3} F_2)\, dx_2 \wedge dx_3 \;+\; (\partial_{x_3} F_1 - \partial_{x_1} F_3)\, dx_3 \wedge dx_1 \;+\; (\partial_{x_1} F_2 - \partial_{x_2} F_1)\, dx_1 \wedge dx_2. \end{align*} The three coefficients are exactly the three components of $\nabla \times F$. Let us write $G = \nabla \times F$ for brevity; then \begin{align*} d\omega_F \;=\; G_1\, dx_2 \wedge dx_3 \;+\; G_2\, dx_3 \wedge dx_1 \;+\; G_3\, dx_1 \wedge dx_2. \end{align*} This combination — vector field component paired with the "opposite" wedge of coordinate $1$-forms — is the **flux $2$-form** of the vector field $G$ on $\mathbb{R}^3$. It can also be written as the Hodge dual $\star_3 G^\flat$ of the corresponding $1$-form, but we will not need that formalism here. This is the well-known statement that **the [exterior derivative](/theorems/1525) on $1$-forms in $\mathbb{R}^3$ reproduces the curl**, recorded as [Exterior Derivative Recovers Gradient, Curl, and Divergence on $\mathbb{R}^3$](/theorems/3566). The miracle hand-wave is avoided: every term came from the antisymmetric wedge. [/guided] [/step] [step:Identify the flux integral $\int_S (\nabla \times F)\cdot n \, d\mathcal{H}^2$ with $\int_S d\omega_F$] Let $p \in S$ and let $(e_1, e_2)$ be any positively-oriented orthonormal basis of $T_pS \subseteq T_p\mathbb{R}^3 \cong \mathbb{R}^3$, so that $(e_1, e_2, n(p))$ is a positively-oriented orthonormal basis of $\mathbb{R}^3$. By the [Integration of Differential Forms](/theorems/1529) for a compactly supported $2$-form on an oriented $2$-manifold, \begin{align*} \int_S d\omega_F \;=\; \int_S (d\omega_F)_p(e_1, e_2) \, d\mathcal{H}^2(p), \end{align*} where the integrand is evaluated at each $p \in S$ on any positively-oriented orthonormal frame of $T_pS$ (the value does not depend on the choice, by the alternating-multilinear property). Fix $p \in S$ and set $G(p) := (\nabla \times F)(p)$. Using the computation of $d\omega_F$ from the previous step, \begin{align*} (d\omega_F)_p(e_1, e_2) \;=\; G_1(p)\, (dx_2 \wedge dx_3)(e_1, e_2) \,+\, G_2(p)\, (dx_3 \wedge dx_1)(e_1, e_2) \,+\, G_3(p)\, (dx_1 \wedge dx_2)(e_1, e_2). \end{align*} For any $u, v \in \mathbb{R}^3$, a direct expansion gives $(dx_i \wedge dx_j)(u, v) = u_i v_j - u_j v_i$, the $(i,j)$-minor of the matrix $[u \, v]$. Equivalently, the triple \begin{align*} \big((dx_2\wedge dx_3)(u,v),\; (dx_3\wedge dx_1)(u,v),\; (dx_1\wedge dx_2)(u,v)\big) \;=\; u \times v. \end{align*} Applying this to $u = e_1$, $v = e_2$: \begin{align*} (d\omega_F)_p(e_1, e_2) \;=\; G(p) \cdot (e_1 \times e_2). \end{align*} Since $(e_1, e_2, n(p))$ is a positively-oriented orthonormal basis of $\mathbb{R}^3$, the cross product satisfies $e_1 \times e_2 = n(p)$. Therefore \begin{align*} (d\omega_F)_p(e_1, e_2) \;=\; G(p) \cdot n(p) \;=\; (\nabla \times F)(p) \cdot n(p), \end{align*} and substituting into the integral definition, \begin{align*} \int_S d\omega_F \;=\; \int_S (\nabla \times F)(p) \cdot n(p) \, d\mathcal{H}^2(p) \;=\; \int_S (\nabla \times F) \cdot n \, d\mathcal{H}^2. \end{align*} [guided] We must show that integrating the $2$-form $d\omega_F$ over the oriented surface $S$ gives the same number as the flux integral $\int_S (\nabla \times F) \cdot n \, d\mathcal{H}^2$. The integral-of-a-$2$-form definition reduces this to a pointwise identity, which we then unwind using linear algebra. **Step A: reduce to a pointwise identity.** Fix $p \in S$. The tangent plane $T_pS$ is a $2$-dimensional subspace of $T_p\mathbb{R}^3 \cong \mathbb{R}^3$, and the orientation of $S$ at $p$ is encoded by the unit normal $n(p)$: a basis $(e_1, e_2)$ of $T_pS$ is **positively oriented** when $(e_1, e_2, n(p))$ is a positively-oriented basis of $\mathbb{R}^3$. Pick any positively-oriented orthonormal basis $(e_1, e_2)$ of $T_pS$. By [Integration of Differential Forms](/theorems/1529) on the oriented Riemannian $2$-manifold $S$, integrating a $2$-form against $\mathcal{H}^2$ is given by \begin{align*} \int_S d\omega_F \;=\; \int_S (d\omega_F)_p(e_1, e_2)\, d\mathcal{H}^2(p), \end{align*} where the integrand is the value of $(d\omega_F)_p$ on the positive orthonormal frame; this value is independent of the chosen orthonormal frame since $\det$ of a rotation in $O(2) \cap SO(T_pS)$ is $+1$ on positively-oriented frames. **Step B: compute the pointwise value.** From the previous step, $d\omega_F = G_1\, dx_2 \wedge dx_3 + G_2\, dx_3 \wedge dx_1 + G_3\, dx_1 \wedge dx_2$ with $G = \nabla \times F$. For each elementary $2$-form, a direct calculation in coordinates shows that for any $u, v \in \mathbb{R}^3$, \begin{align*} (dx_i \wedge dx_j)(u, v) \;=\; \det\begin{pmatrix} u_i & v_i \\ u_j & v_j \end{pmatrix} \;=\; u_i v_j - u_j v_i. \end{align*} Stacking the three resulting numbers as the components of a vector in $\mathbb{R}^3$: \begin{align*} \big((dx_2 \wedge dx_3)(u,v),\; (dx_3 \wedge dx_1)(u,v),\; (dx_1 \wedge dx_2)(u,v)\big) \;=\; (u_2 v_3 - u_3 v_2,\; u_3 v_1 - u_1 v_3,\; u_1 v_2 - u_2 v_1) \;=\; u \times v. \end{align*} This is the **classical identification** of the flux $2$-form basis with the cross product on $\mathbb{R}^3$. Therefore, evaluating on $(e_1, e_2)$: \begin{align*} (d\omega_F)_p(e_1, e_2) \;=\; G_1(p)\,(e_1 \times e_2)_1 \,+\, G_2(p)\,(e_1 \times e_2)_2 \,+\, G_3(p)\,(e_1 \times e_2)_3 \;=\; G(p) \cdot (e_1 \times e_2). \end{align*} **Step C: use the orientation hypothesis.** Since $(e_1, e_2, n(p))$ is positively-oriented orthonormal in $\mathbb{R}^3$, the right-hand rule gives $e_1 \times e_2 = n(p)$. (If we had chosen the opposite orientation, both the value of the $2$-form and $e_1 \times e_2$ would flip sign together — they encode the same geometric data.) Hence \begin{align*} (d\omega_F)_p(e_1, e_2) \;=\; G(p) \cdot n(p) \;=\; (\nabla \times F)(p) \cdot n(p). \end{align*} Inserting into the integral, \begin{align*} \int_S d\omega_F \;=\; \int_S (\nabla \times F) \cdot n \, d\mathcal{H}^2, \end{align*} which is the left-hand side of the Classical Stokes statement. [/guided] [/step] [step:Apply the Generalised Stokes' Theorem to $\omega_F$ on $S$] We verify the hypotheses of the [Generalised Stokes' Theorem](/theorems/3555). It requires: (i) a compact, oriented, smooth $k$-dimensional manifold-with-boundary $M$, with $\partial M$ given the induced (Stokes) orientation; (ii) a smooth $(k-1)$-form on a neighbourhood of $M$. Here $M = S$ has dimension $k = 2$, is compact, oriented (by $n$), and is a smooth submanifold-with-boundary of $\mathbb{R}^3$ by hypothesis. The boundary $\partial S$ carries the induced Stokes orientation by hypothesis, which is exactly the orientation specified in (i): a unit positive tangent $\tau$ to $\partial S$ together with the outward conormal $\nu \in T_p S$ to $\partial S$ along $S$ satisfies $(\nu, \tau, n)$ positively-oriented in $\mathbb{R}^3$. The form $\omega_F$ is a smooth $1$-form on $U$, an open neighbourhood of $S$, so it is a smooth $(k-1) = 1$-form on a neighbourhood of $M$. Both hypotheses hold. Applying the theorem: \begin{align*} \int_S d\omega_F \;=\; \int_{\partial S} \omega_F. \end{align*} Combining with the identifications established in the previous steps, \begin{align*} \int_S (\nabla \times F) \cdot n \, d\mathcal{H}^2 \;\overset{\text{Step 4}}{=}\; \int_S d\omega_F \;\overset{\text{Gen. Stokes}}{=}\; \int_{\partial S} \omega_F \;\overset{\text{Step 2}}{=}\; \oint_{\partial S} F \cdot \tau \, d\mathcal{H}^1. \end{align*} This is the desired identity, completing the proof. $\blacksquare$ [guided] Everything we have done so far has been **translation**: we rewrote both sides of the Classical Stokes identity in the language of differential forms. The actual analytical content — the equality of the two integrals — is exactly the [Generalised Stokes' Theorem](/theorems/3555). Let us verify its hypotheses. The [Generalised Stokes' Theorem](/theorems/3555) applies to a compact, oriented, smooth $k$-dimensional manifold-with-boundary $M$, with $\partial M$ in its induced (Stokes) orientation, and a smooth $(k-1)$-form defined on a neighbourhood of $M$. We take $M = S$ and $k = 2$. We check each condition: - **$S$ is a smooth compact oriented $2$-manifold-with-boundary.** This is part of our hypothesis: $S$ is given as a compact oriented smooth surface with boundary in $\mathbb{R}^3$, with orientation specified by the continuous unit normal $n$. - **$\partial S$ carries the induced Stokes orientation.** This is also part of our hypothesis. The convention used by [Generalised Stokes' Theorem](/theorems/3555) — that a positive tangent $\tau$ to $\partial S$ paired with the outward conormal $\nu \in T_pS$ satisfies $(\nu, \tau, n)$ positively-oriented in $\mathbb{R}^3$ — is the same convention we stated. - **$\omega_F$ is a smooth $1$-form on a neighbourhood of $S$.** Since $F: U \to \mathbb{R}^3$ has smooth components $F_i \in C^\infty(U)$ on the open neighbourhood $U \supseteq S$, and $\omega_F = \sum F_i\, dx_i$ in standard coordinates, $\omega_F \in \Omega^1(U)$. All three hypotheses hold, and the theorem gives \begin{align*} \int_S d\omega_F \;=\; \int_{\partial S} \omega_F. \end{align*} Now we assemble the chain. From Step 4, the left-hand side of Classical Stokes equals the left-hand side here: \begin{align*} \int_S (\nabla \times F) \cdot n \, d\mathcal{H}^2 \;=\; \int_S d\omega_F. \end{align*} From Step 2, the right-hand side of Generalised Stokes equals the right-hand side of Classical Stokes: \begin{align*} \int_{\partial S} \omega_F \;=\; \oint_{\partial S} F \cdot \tau \, d\mathcal{H}^1. \end{align*} Stringing them together, \begin{align*} \int_S (\nabla \times F) \cdot n \, d\mathcal{H}^2 \;=\; \int_S d\omega_F \;=\; \int_{\partial S} \omega_F \;=\; \oint_{\partial S} F \cdot \tau \, d\mathcal{H}^1, \end{align*} which is the Classical Stokes identity. $\blacksquare$ [/guided] [/step]