1 Definition and basic properties

1.1 Holomorphic and meromorphic functions

Let \(D\subset\mathbb{C}\) be a domain. A function \(f:D\to\widehat{\mathbb{C}}\) (where \(\widehat{\mathbb{C}}=\mathbb{C}\cup\{\infty\}\)) is meromorphic on \(D\) if it is holomorphic on \(D\) except at a set of isolated points, and at each such point the function has a pole (so the value \(\infty\) is realized only at isolated locations). In other words, the singularities are allowed, but only in a controlled manner.

Equivalently, one often defines meromorphic functions as those that are holomorphic to \(\widehat{\mathbb{C}}\) when \(\infty\) is treated as a point of the Riemann sphere. With this viewpoint, the singularities that are not removable appear as poles.

1.2 Isolated singularities

If \(a\in D\) is an isolated singularity of a holomorphic function defined on \(D\setminus\{a\}\), one classifies the local behavior into removable singularities, poles, or essential singularities.

1.2.1 Removable singularities

A singularity at \(a\) is removable if \(f\) remains bounded near \(a\) (equivalently, if the singularity can be “filled in” by defining a suitable value \(f(a)\) so that the extended function becomes holomorphic on all of \(D\)). Removable singularities are the mildest type because they introduce no genuine blow-up.

1.2.2 Poles

A singularity at \(a\) is a pole if \(f(z)\to\infty\) as \(z\to a\), and the divergence has a finite order. Poles are precisely the singularities allowed in the standard notion of meromorphic functions.

1.2.3 Essential singularities

A singularity is essential if it is neither removable nor a pole. In that case the function’s values near the point are highly irregular: the function fails to approach any limit and its behavior cannot be captured by a finite principal part.

1.3 Local representation as a quotient

A fundamental feature of meromorphic functions is their local description. If \(f\) is meromorphic near \(a\), then in a neighborhood there exist holomorphic functions \(g\) and \(h\) such that \[ f=\frac{g}{h}, \] with \(h\) not identically zero and \(h(z)\neq 0\) except possibly at isolated points. When \(h(a)\neq 0\), the point is nonsingular for \(f\). When \(h(a)=0\), the order of vanishing of \(h\) determines whether the singularity becomes removable or a pole.

1.4 Relation to rational functions

On \(\widehat{\mathbb{C}}\), rational functions are meromorphic: their only possible singularities occur at isolated points (including possibly at infinity), and those singularities are poles. Thus rational functions serve as the basic global examples of meromorphic functions on the Riemann sphere.

More generally, meromorphic functions on other complex spaces behave like “rational functions with respect to the geometry” of those spaces, with poles occurring at discrete sets of points on the domain.

2 Examples

2.1 Simple rational functions

Functions of the form \[ f(z)=\frac{P(z)}{Q(z)} \] with polynomials \(P,Q\) and \(Q\not\equiv 0\) are meromorphic on \(\mathbb{C}\). The poles are located at zeros of \(Q\), and the order of each pole matches the order of the corresponding zero of \(Q\) (after cancellation with factors shared by \(P\)).

At infinity, rational functions behave like a quotient of polynomials in \(z\), so the point at infinity is either removable, a pole, or regular depending on the degree comparison of \(P\) and \(Q\).

2.2 The gamma function

The Gamma function \(\Gamma(z)\) is meromorphic on \(\mathbb{C}\). It has poles at the non-positive integers \(0,-1,-2,\dots\), and these poles are simple. Away from these points, \(\Gamma\) is holomorphic, making it a classical example of a meromorphic function with an infinite discrete set of poles.

Meromorphic functions arise naturally from quotients of entire functions. For instance:

  • \(\tan z=\frac{\sin z}{\cos z}\) is meromorphic with poles at \(\frac{\pi}{2}+k\pi\).
  • \(\sec z=\frac{1}{\cos z}\) is meromorphic with the same pole set.
  • Ratios like \(\frac{e^z-1}{z}\) can also be meromorphic, with removable singularities possible at \(z=0\).

Even where poles occur, the local behavior is determined by the orders of zeros of the denominator.

2.4 Elliptic and modular examples

Elliptic functions are meromorphic functions on \(\mathbb{C}\) that are doubly periodic. They have discrete sets of poles within a fundamental parallelogram, and their periodicity forces the poles to repeat on the entire lattice.

Modular forms and related objects often appear through meromorphic functions on modular curves, where the allowed singularities correspond to the behavior at cusps and the geometry of the associated domain.

3 Poles and singular behavior

3.1 Order of a pole

If \(a\) is a pole of \(f\), the order (or multiplicity) of the pole is the smallest positive integer \(m\) such that \[ (z-a)^m f(z) \] extends holomorphically and is nonzero at \(z=a\). Equivalently, \(f\) admits a Laurent expansion with leading term proportional to \((z-a)^{-m}\), and no more singular terms occur.

3.2 Laurent series near a pole

Near an isolated singularity \(a\), any meromorphic function has a Laurent expansion \[ f(z)=\sum_{n=-m}^{\infty} c_n (z-a)^n \] for some \(m\ge 0\), where \(m=0\) corresponds to a removable singularity. When \(a\) is a pole, the expansion contains finitely many negative-power terms, and the smallest exponent determines the order of the pole.

3.3 Behavior near infinity

A meromorphic function on \(\mathbb{C}\) may also be viewed near the point at infinity by studying the transformed function \(g(w)=f(1/w)\) near \(w=0\). Depending on whether \(f\) grows, decays, or stays bounded sufficiently fast, the point at infinity can behave like:

  • a removable singularity,
  • a pole,
  • or a more general singularity in the meromorphic sense on \(\widehat{\mathbb{C}}\).

For rational functions, the classification at infinity is particularly direct in terms of degrees.

3.4 Principal parts

The principal part of \(f\) at a pole \(a\) is the finite sum of negative-power terms from the Laurent expansion: \[ \operatorname{PP}_a(f)(z)=\sum_{n=1}^{m} c_{-n}(z-a)^{-n}. \] It captures the singular contribution, while the remaining series corresponds to a holomorphic function near \(a\).

4 Operations on meromorphic functions

4.1 Addition and subtraction

If \(f\) and \(g\) are meromorphic on the same domain, then \(f\pm g\) is meromorphic. The singular set of \(f\pm g\) is contained in the union of the singular sets of \(f\) and \(g\). At isolated points, cancellations may reduce the order of poles or even remove a pole entirely, but the result remains meromorphic.

4.2 Multiplication and division

Products of meromorphic functions are meromorphic, with pole orders adding when no cancellation occurs. If \(g\) is not identically zero and one forms \(f/g\), the result is meromorphic away from the zeros of \(g\). Zeros of \(g\) become poles of \(f/g\) unless offset by corresponding zeros of \(f\).

4.3 Composition

If \(f\) is meromorphic on a domain \(D\) and \(\phi\) is holomorphic (or meromorphic in compatible settings) mapping into a region where \(f\) is defined, then \(f\circ \phi\) is meromorphic wherever the composition makes sense. Poles occur at points where \(\phi\) hits poles of \(f\), provided the local structure does not collapse under cancellation.

4.4 Differentiation and integration

Differentiation preserves meromorphicity: the derivative of a meromorphic function is meromorphic, with pole order typically increasing by at most one at each pole.

Integration requires a choice of path when defining antiderivatives globally, but locally one can integrate the Laurent series term by term. A meromorphic function admits a local primitive whose behavior near poles is controlled: integrating can turn simple poles into logarithmic terms, though within the meromorphic framework one usually considers derivatives and local primitives carefully in contexts where single-valuedness is ensured.

5 Laurent series and residues

5.1 Laurent expansions

The Laurent expansion is the primary analytic tool for meromorphic functions near isolated singularities. For \(a\) in the domain, write \[ f(z)=\sum_{n=-\infty}^{\infty} c_n (z-a)^n \] in an annulus where the expansion is valid. Meromorphicity implies that for poles the negative part is finite.

5.2 Residue at a pole

The residue of \(f\) at \(a\) is the coefficient of \((z-a)^{-1}\) in the Laurent expansion: \[ \operatorname{Res}(f,a)=c_{-1}. \] Residues encode how contour integrals capture local singular behavior. For poles of order \(m\), residues can be computed using derivative formulas derived from the Laurent coefficients.

5.3 Residue theorem

A central result links sums of residues to integrals. If \(f\) is meromorphic in a region containing a closed contour \(\Gamma\) and finitely many singularities inside \(\Gamma\), then \[ \frac{1}{2\pi i}\int_{\Gamma} f(z)\,dz=\sum_{a\in \text{inside }\Gamma}\operatorname{Res}(f,a), \] counting each singularity once with respect to its position relative to the contour orientation. This theorem turns complex contour integrals into algebraic data.

5.4 Applications to contour integration

Residues are frequently used to evaluate integrals that would be difficult by direct real-variable methods. By selecting contours that exploit symmetry or decay, one can transform real integrals into sums of residues at poles, yielding closed forms or efficient asymptotic behavior.

6 Zeros and poles

6.1 Order of zeros

If \(f\) is holomorphic and \(f(a)=0\), the order of the zero is the largest integer \(k\) such that \[ f(z)=(z-a)^k u(z), \] where \(u\) is holomorphic and \(u(a)\neq 0\). For meromorphic functions, zeros are defined at points where the function takes the value \(0\), and their order is similarly determined by the local factorization.

6.2 Counting zeros and poles

Local orders provide a way to count behavior in neighborhoods. On a compact Riemann surface, global counting becomes structured: the total number of zeros and poles of a meromorphic function, counted with multiplicity, obeys constraints determined by the surface’s topology and the function’s divisor.

In the plane, one can often relate counts inside a region to contour integrals, especially when combined with the residue theorem applied to logarithmic derivatives.

6.3 Argument principle

For a meromorphic function \(f\) that is holomorphic on and near a contour \(\Gamma\) except for isolated poles inside, the argument principle relates the change in argument of \(f\) along \(\Gamma\) to the number of zeros and poles inside. In its common form, it expresses this change in terms of an integral of \(f'(z)/f(z)\), which is naturally suited to meromorphic functions because the logarithmic derivative has poles at zeros and poles of \(f\).

6.4 Divisors of meromorphic functions

A divisor records zeros and poles with multiplicities. For a meromorphic function \(f\) on a Riemann surface, its divisor is a formal sum of points weighted by the orders of \(f\) at those points (positive for zeros, negative for poles). This organizes meromorphic data into an algebraic structure that supports further classification and comparison of functions.

7 Analytic continuation

7.1 Extending meromorphic functions

Meromorphic functions often arise initially on a small region and can be extended by analytic continuation. If two meromorphic functions agree on a set with an accumulation point, they coincide on the connected component of their common domain of definition. This uniqueness principle ensures that once a continuation exists, it is essentially determined.

Continuation may fail if the function encounters natural barriers tied to singularities, but in many settings one can enlarge the domain while preserving meromorphic behavior.

7.2 Branch points versus poles

Not all singularities in continuation are poles. When analytic continuation follows a path around certain singularities, the function might return to a different value or different branch, indicating branch points. In contrast, poles correspond to meromorphic behavior: they remain single-valued and their singularity type does not produce multivaluedness.

Distinguishing these cases is important when classifying the maximal extension of a function.

7.3 Maximal meromorphic continuation

A maximal meromorphic continuation is the largest domain on which a function extends as a single-valued meromorphic function, consistent with its analytic continuation from the initial data. In many classical examples, continuation leads to a global meromorphic function on a compact surface or on the Riemann sphere, whereas in others one must pass to a covering space to handle branching.

8 Global theory on complex domains

8.1 Meromorphic functions on open sets

On an open set \(D\subset\mathbb{C}\), meromorphic functions form a natural class closed under the operations typically performed in complex analysis. Their singular sets are discrete, so they behave like holomorphic functions except at isolated poles. This local-to-global viewpoint underlies most practical uses of meromorphic functions on planar domains.

8.2 Meromorphic functions on the Riemann sphere

On \(\widehat{\mathbb{C}}\), meromorphic functions are exactly the rational functions. This characterization reflects the compactness of the sphere and the strong constraints it imposes: singularities must be poles, and the global structure forces a finite-degree description.

8.3 Meromorphic functions on compact Riemann surfaces

On a compact Riemann surface \(X\), meromorphic functions are global objects defined by holomorphic maps into \(\widehat{\mathbb{C}}\). Poles occur at finitely many points on \(X\), and meromorphic functions become central in the classification of maps from \(X\) to the sphere.

8.3.1 Rational functions on the sphere

On the sphere, the theory specializes to classical rational function behavior. Poles are determined by the points where the denominator vanishes, and divisors encode degree data. This case serves as the prototype for more general compact surfaces.

8.3.2 Meromorphic functions on tori

A torus can be realized as \(\mathbb{C}/\Lambda\) for a lattice \(\Lambda\). Meromorphic functions on the torus are periodic with respect to the lattice and have finitely many poles per fundamental region. Their structure is often studied through elliptic functions, which provide explicit examples and tools.

8.3.3 Abelian functions

In broader contexts, abelian functions describe meromorphic functions on complex tori of higher dimension or related algebraic varieties, often connected to the theory of theta functions and integrals. While the explicit constructions differ, the unifying theme remains: controlled singularities and global meromorphic behavior governed by the underlying geometry.

9 Connections and applications

9.1 Complex differential equations

Many linear ordinary differential equations with analytic coefficients yield solutions expressible in terms of meromorphic functions or functions closely related to them. Poles and their orders often correspond to local singular behaviors dictated by the differential equation’s structure. In this way, meromorphic functions provide a language for both existence results and asymptotic or local classification.

9.2 Number theory

Meromorphic functions appear throughout number theory, notably in generating functions and in complex-analytic frameworks for arithmetic objects. Their poles and residues can correspond to coefficients, special values, or transformation properties. The controlled nature of meromorphic singularities makes them suitable for extracting global arithmetic information from analytic data.

9.3 Complex geometry

In complex geometry, meromorphic functions act as mappings to \(\widehat{\mathbb{C}}\) and serve as basic building blocks for understanding divisors, line bundles, and morphisms. The correspondence between divisors and meromorphic behavior links analytic properties to geometric invariants, enabling classification and intersection-theoretic interpretations.

9.4 Mathematical physics

Meromorphic functions arise in mathematical physics via complex potentials, Green’s functions, spectral resolvents, and other objects whose singularities represent physical features such as resonances or interactions localized in space. Residue calculus and contour integration, powered by the meromorphic framework, often provide systematic ways to evaluate integrals and derive relationships between physical quantities.