Cauchy's Integral Formula

Cauchy's Integral Formula is a cornerstone of complex analysis that expresses the value of a holomorphic (complex-differentiable) function $f$ at any point $a$ inside a simple closed contour $C$ as a weighted average of its boundary values. Precisely, if $f$ is holomorphic on and inside a positively oriented simple closed curve $C$, then $f(a)$ is recovered by integrating $f(z)/(z-a)$ around $C$. This matters because it transforms local information (a single value) into a statement about global boundary data, and it leads directly to the astonishing fact that every holomorphic function is infinitely differentiable and analytic — representable by a convergent power series. The surprising fact: knowing a holomorphic function on the boundary of a region fixes it everywhere inside, with no freedom whatsoever — a rigidity that has no analogue for real differentiable functions.

The Core Insight

A holomorphic function is so tightly constrained that its boundary values broadcast inward, determining everything in the interior. The kernel $\frac{1}{z-a}$ acts like a "sampling probe" that, when integrated around a loop, sifts out exactly the value $f(a)$. This single mechanism is responsible for analyticity, the maximum modulus principle, Liouville's theorem, and the residue calculus.

Formal Statement

Holomorphic Function and Contour

A function $f:\Omega \to \mathbb{C}$ defined on an open set $\Omega \subseteq \mathbb{C}$ is holomorphic at $z_0$ if the limit

exists, where $h\in \mathbb{C}$. A contour (or closed curve) $C$ is a piecewise smooth, continuous path $\gamma :[a,b]\to \mathbb{C}$; it is simple closed if $\gamma (a)=\gamma (b)$ and $\gamma$ is injective on $(a,b)$. It is positively oriented if traversed counterclockwise, keeping the enclosed region on the left.

Cauchy's Integral Formula

Let $f$ be holomorphic on an open set containing the closed region bounded by a positively oriented simple closed contour $C$. Then for any point $a$ in the interior of $C$,

More generally, $f$ has derivatives of all orders, given by the generalized Cauchy formula:

Here ${\oint }_C$ denotes the line integral around $C$, $i=\sqrt{-1}$, and $2\pi i$ is the value of ${\oint }_C\frac{dz}{z-a}$ for any simple loop enclosing $a$ once.

Why It Holds

Derivation of the Integral Formula

We prove [2] in numbered steps.

Since $f$ is holomorphic, $g$ has a removable singularity at $a$ because ${\lim }_{z\to a}g(z)=f^{\prime }(a)$ exists and is finite; thus $g$ extends to a holomorphic function on the whole region.

Sketch for the Derivative Formula

To obtain [3], write the difference quotient of [2]:

As $h\to 0$ the bracket tends to $\frac{1}{(z-a{)}^2}$ uniformly on $C$ (which is bounded away from $a$), so passage to the limit under the integral gives $f^{\prime }(a)=\frac{1}{2\pi i}{\oint }_C\frac{f(z)}{(z-a{)}^2}dz$. Induction on $n$, repeating this argument, produces the general formula [3]. $■$

Augustin-Louis Cauchy | 1789–1857

Cauchy was a prolific French mathematician who, almost single-handedly, made analysis rigorous. Working in Paris during a turbulent era of revolution and restoration, he introduced precise definitions of limits, continuity, and convergence, and founded the modern theory of functions of a complex variable. His 1825 memoir on definite integrals between complex limits and his 1831 work in Turin gave the integral theorem and integral formula that bear his name. Famously devout and politically royalist, Cauchy refused oaths of allegiance and went into self-exile, yet never stopped publishing — producing nearly 800 papers across analysis, elasticity, optics, and celestial mechanics.

Building Intuition

Picture the kernel $\frac{1}{z-a}$ as a probe that "tunes in" to the point $a$. As $z$ travels once around the loop $C$, the factor $\frac{1}{z-a}$ winds once around the origin, contributing the universal phase factor $2\pi i$. Multiplying by $f(z)$ and integrating averages the boundary values, but the winding singularity at $a$ pulls the average toward $f(a)$ exactly. Another picture: holomorphic functions behave like incompressible, irrotational flows; their boundary behavior rigidly determines the interior, just as the boundary velocity of an ideal fluid determines the entire flow pattern. This is why analytic continuation works — once you know $f$ on any small piece, the formula forces its values everywhere accessible.

Contour C enclosing the point a, with deformation to a small circle

Worked Examples

Evaluating a Contour Integral Directly

1. Identify $f(z)=e^z$ and the singularity $a=2$.
2. Check that $a=2$ lies inside $C$: since $|2|=2<3$, yes.
3. Confirm $f(z)=e^z$ is holomorphic everywhere (entire), so the hypotheses of [2] hold.
4. Apply Cauchy's Integral Formula [2] rearranged: ${\oint }_C\frac{f(z)}{z-a}dz=2\pi if(a)$.
5. Compute $f(2)=e^2$.

Using the Derivative Formula

1. Match to formula [3] with $f(z)=\cos z$, $a=0$, and exponent $n+1=3$, so $n=2$.
2. The formula reads ${\oint }_C\frac{f(z)}{(z-a{)}^3}dz=\frac{2\pi i}{2!}{f}^{\prime \prime }(a)$.
3. Differentiate: $f^{\prime }(z)=-\sin z$, $f^{\prime \prime }(z)=-\cos z$.
4. Evaluate $f^{\prime \prime }(0)=-\cos 0=-1$.
5. Insert: $\frac{2\pi i}{2}(-1)=-\pi i$.

Common Misconceptions

Consider $f(z)=\bar{z}$ (complex conjugate). One might hope to apply the formula, but $f$ is not holomorphic anywhere — it fails the Cauchy–Riemann equations. Indeed ${\oint }_{|z|=1}\frac{\bar{z}}{z}dz$. On $|z|=1$ we have $\bar{z}=1/z$, so the integrand is $1/z^2$, and ${\oint }_{|z|=1}\frac{dz}{z^2}=0$, which does not equal $2\pi i\bar{a}=2\pi i\cdot 0=0$ by coincidence here — but for general $a$ inside, $f(a)=\bar{a}$ would be required, which the integral cannot reproduce. The formula simply does not apply to non-holomorphic functions; matching values is accidental and breaks down immediately for other contours or points.

Connections

Cauchy's Integral Formula sits at the hub of complex analysis. It is an immediate consequence of Cauchy's Integral Theorem (the vanishing of integrals of holomorphic functions over closed loops). Its derivative form [3] proves that holomorphic functions are infinitely differentiable, hence equal to their Taylor Series — the principle of Analyticity of Holomorphic Functions. Extracting the coefficient of $(z-a{)}^{-1}$ generalizes the formula into the Residue Theorem, the master tool for evaluating real integrals. It directly yields Liouville's Theorem (bounded entire functions are constant), the Maximum Modulus Principle, and the Fundamental Theorem of Algebra. The mean value form (taking $C$ a circle) connects to Harmonic Functions and the Poisson integral.

Real-World Applications

1. Evaluation of real definite integrals — integrals like ${\int }_{-\infty }^{\infty }\frac{dx}{1+x^2}$ are computed by closing a contour in the complex plane and applying the formula/residues, yielding $\pi$.
2. Signal processing & control theory — the Nyquist stability criterion counts encirclements of points via contour integrals, directly built on the argument principle descended from this formula.
3. Fluid dynamics — two-dimensional ideal (potential) flow is described by a holomorphic complex potential; the formula recovers velocity fields from boundary data, used in airfoil theory (Joukowski transformations).
4. Electrostatics — the complex potential for 2D charge distributions is holomorphic; field values inside a region follow from boundary potentials.
5. Quantum field theory & dispersion relations — the Kramers–Kronig relations linking the real and imaginary parts of response functions are derived using Cauchy's formula on the upper half-plane.
6. Numerical analysis — high-accuracy computation of derivatives via the Cauchy integral method (Lyness–Moler) avoids subtractive cancellation by integrating around a circle.
7. String theory & conformal field theory — operator product expansions and correlation functions are organized by contour-integral (radial-ordering) techniques rooted in Cauchy's formula.

Historical Development

• 1814 — Cauchy presents an early memoir on definite integrals, planting the seeds of complex integration while studying real integrals via complex methods.
• 1822–1825 — Cauchy develops the integral theorem (closed-contour integrals of holomorphic functions vanish), published in his 1825 Mémoire sur les intégrales définies prises entre des limites imaginaires.
• 1831 — While in exile in Turin, Cauchy states the integral formula and his "calcul des limites" (later the Cauchy estimates), giving the power-series representation of analytic functions.
• 1843–1846 — Cauchy systematizes the residue theorem, generalizing the integral formula to functions with poles.
• 1851 — Bernhard Riemann's dissertation recasts complex functions geometrically via the Cauchy–Riemann equations and Riemann surfaces, deepening the conceptual foundation.
• 1860s — Karl Weierstrass develops the rigorous power-series ("analytic element") approach, providing an alternative foundation independent of integration.
• 1883 — Édouard Goursat proves the integral theorem assuming only differentiability (not continuous derivatives), removing a hidden hypothesis and yielding the modern Cauchy–Goursat theorem.
• 1900s — The homological/winding-number formulation (via the index of a curve) generalizes the formula to arbitrary cycles in multiply connected regions.

Subtleties & Limits

The classical formula assumes $f$ holomorphic on a simply connected region. Subtleties arise in multiply connected domains, where the general statement uses the winding number $n(C,a)=\frac{1}{2\pi i}{\oint }_C\frac{dz}{z-a}$, giving $n(C,a)f(a)=\frac{1}{2\pi i}{\oint }_C\frac{f(z)}{z-a}dz$ — the loop may encircle $a$ several times or not at all. Extending integral-representation ideas to several complex variables leads to the Bochner–Martinelli and Cauchy–Fantappiè kernels, where the geometry is vastly richer and many natural questions (e.g., explicit reproducing kernels on general pseudoconvex domains, behavior of the $\bar{\partial }$-Neumann problem) remain active research. In a different direction, Cauchy-type integrals with rough boundary data connect to deep open questions in harmonic analysis (the Cauchy transform on non-smooth curves, related to the Calderón program), some of which were only resolved late in the 20th century and whose higher-dimensional analogues persist.

See Also