16 Outlook

The four integral theorems — the fundamental theorem for line integrals, Green’s theorem, Stokes’ theorem, and the Divergence theorem — were proved separately in this course, each by its own argument adapted to its own geometry. But they share an unmistakable structure: an integral over a domain equals an integral of a derivative over its boundary. By the end of Chapter 12 it was clear that this is not a coincidence. It is one theorem, seen in four different dimensions.

The language that makes this precise is the theory of differential forms. This chapter introduces that language without proofs — the proofs belong in Lee’s Introduction to Smooth Manifolds or Spivak’s Calculus on Manifolds, both of which pick up almost exactly where this course ends. Think of what follows as a map of the territory immediately adjacent to what we have covered, drawn to give you a sense of what is there and how to reach it.

16.1 Differential Forms

Throughout this course, the integrands in our line and surface integrals involved expressions like P\,dx + Q\,dy or F_1\,dy\wedge dz + F_2\,dz\wedge dx + F_3\,dx\wedge dy. These are differential forms — we have been using them all along without the name. The correct framework organises them by degree and makes parametrisation-independence automatic.

The starting point is the wedge product: an antisymmetric bilinear operation on differentials satisfying dx^i \wedge dx^j = -dx^j \wedge dx^i \qquad \text{for all } i,j. In particular dx^i \wedge dx^i = 0. The wedge product is associative and distributes over addition. It is antisymmetric precisely in the sense that encodes orientation: swapping two differentials changes the sign.

Definition 16.1 (Differential k-form) A differential k-form on an open set U \subset \mathbb{R}^n is an expression \omega = \sum_{i_1 < i_2 < \cdots < i_k} f_{i_1 \cdots i_k}\, dx^{i_1} \wedge dx^{i_2} \wedge \cdots \wedge dx^{i_k}, where the f_{i_1\cdots i_k} : U \to \mathbb{R} are smooth functions. The space of smooth k-forms on U is denoted \Omega^k(U).

The integer k is the degree of the form. A k-form can be integrated over an oriented k-dimensional submanifold of U.

The cases we have encountered:

A 0-form is just a smooth function f \in \Omega^0(U) = C^\infty(U).

A 1-form is \omega = \sum_i f_i\, dx^i \in \Omega^1(U). On \mathbb{R}^3, \omega = P\,dx + Q\,dy + R\,dz. Its integral over a curve \gamma is the line integral \int_\gamma \omega = \int_\gamma F \cdot d\gamma with F = (P,Q,R).

A 2-form on \mathbb{R}^3 is \omega = f\,dy\wedge dz + g\,dz\wedge dx + h\,dx\wedge dy \in \Omega^2(\mathbb{R}^3). Its integral over an oriented surface S is the flux integral \iint_S \omega = \iint_S F \cdot d\mathbf{S} with F = (f,g,h). The antisymmetry of \wedge encodes orientation: reversing the order of dx and dy negates the form, which is the same as reversing the surface normal.

A 3-form on \mathbb{R}^3 is \omega = f\,dx\wedge dy\wedge dz. Its integral over a region \Omega is \iiint_\Omega \omega = \iiint_\Omega f\,dV.

There are no nonzero k-forms on \mathbb{R}^n for k > n, since dx^i \wedge dx^i = 0 forces any such form to be zero.

16.2 The Exterior Derivative

The \nabla \to \operatorname{curl} \to \operatorname{div} chain from Chapter 5 is the shadow of a single operation on differential forms.

Definition 16.2 (Exterior derivative) The exterior derivative is the linear map d : \Omega^k(U) \to \Omega^{k+1}(U) defined on a k-form \omega = \sum_{I} f_I\, dx^I (where I = (i_1 < \cdots < i_k) is a multi-index and dx^I = dx^{i_1} \wedge \cdots \wedge dx^{i_k}) by d\omega = \sum_I df_I \wedge dx^I = \sum_I \sum_{j=1}^n \frac{\partial f_I}{\partial x^j}\, dx^j \wedge dx^I.

On the three cases in \mathbb{R}^3:

d applied to a 0-form f gives df = \partial_1 f\,dx + \partial_2 f\,dy + \partial_3 f\,dz — the gradient, written as a 1-form.
d applied to the 1-form \omega = P\,dx + Q\,dy + R\,dz gives the 2-form with coefficients (\partial_y R - \partial_z Q,\, \partial_z P - \partial_x R,\, \partial_x Q - \partial_y P) — the curl.
d applied to the 2-form \omega = f\,dy\wedge dz + g\,dz\wedge dx + h\,dx\wedge dy gives (\partial_x f + \partial_y g + \partial_z h)\, dx\wedge dy\wedge dz — the divergence.

The two identities \operatorname{curl}(\nabla f) = 0 and \operatorname{div}(\operatorname{curl} F) = 0 (Theorem 7.2) from Chapter 5 both follow from a single algebraic fact:

Theorem 16.1 (d^2 = 0) For any smooth form \omega, d(d\omega) = 0.

Proof deferred. See Spivak, Calculus on Manifolds, or Lee, Introduction to Smooth Manifolds. \square

The chain \Omega^0(U) \xrightarrow{d} \Omega^1(U) \xrightarrow{d} \Omega^2(U) \xrightarrow{d} \Omega^3(U) in which every consecutive composition is zero is called the de Rham complex of U. Whether every closed form (d\omega = 0) is exact (\omega = d\eta for some \eta) depends on the topology of U — this is the subject of de Rham cohomology, developed in Lee and in Bott–Tu Differential Forms in Algebraic Topology.

16.3 The General Stokes’ Theorem

With this language, all four integral theorems collapse into one.

Theorem 16.2 (Generalised Stokes’ theorem) Let M be a compact oriented smooth k-manifold with boundary \partial M (an oriented (k-1)-manifold), and let \omega \in \Omega^{k-1}(U) for an open set U \supset M. Then \int_{\partial M} \omega = \int_M d\omega.

Proof deferred. See Spivak, Calculus on Manifolds, or Lee, Introduction to Smooth Manifolds. \square

This single equation contains every integral theorem in this course:

Fundamental theorem of calculus (k=1, M = [a,b], \omega = f a 0-form, \partial M = \{a,b\} with orientation +b-a): reads f(b) - f(a) = \int_a^b df.

Fundamental theorem for line integrals (k=1, M a curve, \omega = f a 0-form): reads f(\gamma(b)) - f(\gamma(a)) = \int_\gamma df = \int_\gamma \nabla f \cdot d\gamma.

Green’s theorem (k=2, M = \Omega \subset \mathbb{R}^2, \omega = P\,dx + Q\,dy): d\omega = (\partial_x Q - \partial_y P)\,dx\wedge dy, reads \oint_{\partial\Omega} P\,dx + Q\,dy = \iint_\Omega (\partial_x Q - \partial_y P)\,dA.

Stokes’ theorem (k=2, M = S a surface in \mathbb{R}^3, \omega a 1-form): reads \oint_{\partial S} F\cdot d\gamma = \iint_S \operatorname{curl} F\cdot d\mathbf{S}.

Divergence theorem (k=3, M = \Omega \subset \mathbb{R}^3, \omega a 2-form): reads \oiint_{\partial\Omega} F\cdot d\mathbf{S} = \iiint_\Omega \operatorname{div} F\,dV.

The mechanism is the same in all five cases: differentiation of the integrand and integration over the boundary are inverse operations.

16.4 Further Directions

What follows is a map of the natural next courses, with prerequisites from this text listed explicitly.

16.4.1 Ordinary Differential Equations

The study of equations \dot{x} = F(x,t) for x : \mathbb{R} \to \mathbb{R}^n. The existence and uniqueness theorem (Picard–Lindelöf) uses the contraction mapping principle from Chapter 7 — it is perhaps the most direct application of that theorem. The theory of linear systems uses the matrix exponential, eigenvalues, and the spectral theorem.

Prerequisites from this course. The total derivative and chain rule (Chapter 3), the contraction mapping principle and mean value inequality (Chapter 7).

Natural entry point. Immediately after this course.

16.4.2 Partial Differential Equations

The study of equations involving functions of several variables and their partial derivatives: the heat equation \partial_t u = \kappa\Delta u, the wave equation \partial_{tt} u = c^2 \Delta u, Laplace’s equation \Delta u = 0. The Divergence theorem and Green’s identities from Chapter 12 are used constantly; the harmonic function theory of Chapter 13 is the starting point for the elliptic theory.

Prerequisites from this course. Chapters 3–5 (differentiation and vector calculus), Chapters 10–12 (integral theorems), Chapter 13 (harmonic functions and Green’s identities). A course on ODEs first is strongly recommended.

What you will encounter. Separation of variables, Fourier series, eigenfunction expansions. The modern approach via Sobolev spaces requires measure theory and functional analysis.

16.4.3 Real Analysis and Measure Theory

A rigorous treatment of limits, continuity, and integration. This course used the Riemann integral throughout and noted in Chapter 8 where it falls short: it cannot exchange limits and integrals without uniform convergence, and it cannot handle the characteristic function of the rationals. Lebesgue’s theory fixes both. Measurable sets, the Lebesgue integral, and the dominated convergence theorem give the exchange of limits and integrals that the Riemann theory cannot cleanly provide.

Prerequisites from this course. Chapter 1 entirely — topological spaces, metric spaces, compactness, sequences. The epsilon-delta arguments throughout are good preparation for the rigour required.

What you will encounter. L^p spaces, product measures (giving the rigorous proof of Fubini), the Radon–Nikodym theorem, and the measure- theoretic foundations of probability. This is a prerequisite for functional analysis, advanced probability, and rigorous PDE theory.

16.4.4 Point-Set Topology

The systematic study of topological spaces in full generality, without any metric structure. Chapter 1 gave a working introduction in \mathbb{R}^n; point-set topology develops the same ideas — continuity, compactness, connectedness — for arbitrary topological spaces.

Prerequisites from this course. Chapter 1 entirely.

What you will encounter. Urysohn’s lemma, Tychonoff’s theorem (the product of compact spaces is compact), metrisation theorems, and the foundations of algebraic topology. This is a prerequisite for differential geometry and algebraic topology.

16.4.5 Differential Geometry

The study of smooth manifolds with geometric structure: Riemannian metrics, curvature, geodesics, and differential forms in full generality. This course built the foundations — smooth manifolds and tangent spaces in Chapters 1.5 and 2, the derivative as a linear map between tangent spaces in Chapters 3 and 7, and the integral theorems culminating in the generalised Stokes’ theorem above. Differential geometry is the natural continuation.

Prerequisites from this course. Chapters 1.5 and 2 (manifolds, tangent spaces, smooth maps), Chapters 3–7 (differentiation), and the integral theorems of Part 4. Linear algebra — inner product spaces, the spectral theorem — is used throughout.

What you will encounter. Riemannian metrics, connections and parallel transport, the Riemann curvature tensor, and the Gauss–Bonnet theorem — which says the total curvature of a closed surface equals 2\pi times its Euler characteristic, a purely topological number. This is the prototype of the deep connections between geometry and topology that are the subject of modern mathematics.

Standard references. Lee, Introduction to Smooth Manifolds (2nd ed.); do Carmo, Differential Geometry of Curves and Surfaces (for the classical case of surfaces in \mathbb{R}^3 first); Spivak, A Comprehensive Introduction to Differential Geometry.