Mittag-Leffler's Theorem and Meromorphic Functions

Mittag-Leffler's Theorem

Motivation: Infinite Sum of Fractions

A rational function $R(z) = P(z)/Q(z)$ is decomposed into a sum of simple fractions: $R = \sum A_k/(z - z_k)^{m_k} +$ a polynomial. Mittag-Leffler's theorem is an analogue for meromorphic functions with infinitely many poles: one can "prescribe" the poles and their principal parts and construct a function with these data.

Mittag-Leffler's Theorem

Theorem (Mittag-Leffler, 1877): Let ${a_n}$ be a sequence of points without an accumulation point in $\mathbb{C}$ (i.e., $|a_n| \to \infty$), and for each $a_n$ a "principal part" $p_n\left(1/(z - a_n)\right)$ is given—a polynomial in $1/(z - a_n)$. Then there exists a meromorphic function $f$ with poles at ${a_n}$ and the given principal parts:

$ f(z) = h(z) + \sum_n \left[ p_n \left( \frac{1}{z - a_n} \right) - q_n(z) \right], $

where $h(z)$ is an arbitrary entire function, and $q_n(z)$ are polynomials ensuring the convergence of the series.

Partial Fractions of Special Functions

Expansion of $\pi \cot(\pi z)$:

$ \pi \cot(\pi z) = \frac{1}{z} + \sum_{n \ne 0} \left( \frac{1}{z-n} + \frac{1}{n} \right) = \lim_{N \to \infty} \sum_{n = -N}^{N} \frac{1}{z - n}. $

Poles at $z = n \in \mathbb{Z}$, principal parts $1/(z - n)$—simple poles with residue $1$.

From the expansion of $\pi \cot(\pi z)$, integrating yields Euler's product $\sin(\pi z) = \pi z \cdot \prod (1 - z^2/n^2)$.

Expansion of $1/\sin(\pi z)$:

$ \frac{1}{\sin(\pi z)} = \sum_{n = -\infty}^{\infty} \frac{(-1)^n}{z-n}. $

Poles at all integers $n$, residue at $n$ is $(-1)^n$.

Application: Summation of Series

$ \oint f(z) \cdot \pi \cot(\pi z) dz = -2\pi i \sum_{n \in \mathbb{Z}} f(n) $

(the contour encloses all integers $n$).

If $f(z)$ decreases sufficiently rapidly, and the contour is expanded, the integral $\to 0$:

$ \sum_{n = -\infty}^{\infty} f(n) = -\sum_{\text{poles of } f} \operatorname{Res} \left[f(z) \cdot \pi \cot(\pi z) \right]. $

Numerical Example

Problem: Find $\sum_{n=1}^{\infty} 1/(n^2 + a^2)$ via residues.

Step 1. Consider $g(z) = 1/(z^2 + a^2)$ with poles at $z = \pm ia$ (for $a > 0,~a \notin \mathbb{Z}$).

Step 2. $ \sum_{n = -\infty}^{\infty} \frac{1}{n^2 + a^2} = -\operatorname{Res}{z = ia} \left[ \frac{\pi \cot(\pi z)}{z^2 + a^2} \right] - \operatorname{Res}{z = -ia} \left[ \frac{\pi \cot(\pi z)}{z^2 + a^2} \right]. $

Step 3. At $z = ia$: $(z^2 + a^2) = (z - ia)(z + ia)$, residue $= \pi \cot(\pi ia)/(2ia)$.

$\cot(\pi ia) = \frac{\cos(\pi ia)}{\sin(\pi ia)} = \frac{\cosh(\pi a)}{i \sinh(\pi a)} = -i, \coth(\pi a)$.

Residue at $ia$: $\pi \cdot (-i \coth(\pi a))/(2ia) = \pi \coth(\pi a)/(2a)$.

At $z = -ia$: similarly, the sum of the two residues $= \pi \coth(\pi a)/a$.

Step 4. $ \sum_{n = -\infty}^{\infty} \frac{1}{n^2 + a^2} = \frac{\pi \coth(\pi a)}{a}. $

Symmetry: $\frac{1}{a^2} + 2 \cdot \sum_{n = 1}^{\infty} \frac{1}{n^2 + a^2} = \frac{\pi \coth(\pi a)}{a}$.

Result:

$ \sum_{n=1}^{\infty} \frac{1}{n^2 + a^2} = \frac{\pi \coth(\pi a)/a - 1/a^2}{2} = \frac{\pi a \coth(\pi a) - 1}{2 a^2}. $

Check as $a \to 0$: $\coth(\pi a) \approx 1/(\pi a) + \pi a/3 \implies \pi a \coth(\pi a) \approx 1 + \pi^2 a^2/3 \implies \Sigma \approx \pi^2/6 = \sum 1/n^2~\checkmark$.

Real-Life Application

Electrodynamics: image forces in capacitors with multiple plates are computed as sums of the contributions of all "images"—an infinite series, summable via the Mittag-Leffler expansion.

Additional Aspects

The Mittag-Leffler theorem asserts that for any sequence of points $z_n \to \infty$ and prescribed principal parts $g_n(z)$ at those points, there exists a meromorphic function having exactly these poles and principal parts. This is analogous to Weierstrass's theorem for zeros and is widely used in the construction of special functions. For example, the expansion $\pi \cot(\pi z) = 1/z + \sum_{n \geq 1} 2z/(z^2 - n^2)$ gives an elegant proof of Euler's identity $\sum 1/n^2 = \pi^2/6$.

In the theory of differential equations, meromorphic functions describe the resolvents of elliptic operators; their poles correspond to eigenvalues and residues to projectors onto eigenspaces. This underlies spectral methods in numerical linear algebra.

Connection with Other Branches of Mathematics

The Mittag-Leffler theorem naturally complements the Weierstrass theorem on representing entire functions by their zeros; together, they yield the classical theorem on the unique meromorphic function with prescribed zeros and poles (via the logarithmic derivative). This directly connects it to divisors on Riemann surfaces: in the formulation of L. Ahlfors and L. Sario, the problem of constructing a meromorphic function with a given divisor reduces to the global version of the Mittag-Leffler theorem on compact surfaces.

In the theory of differential equations, Mittag-Leffler's theorem is used when analyzing solutions to linear Fuchsian equations and Painlevé equations: according to R. Nevanlinna and E. L. Ingelis, local expansions near singular points are "glued together" globally precisely through such representations. In spectral theory of self-adjoint operators, Mittag-Leffler-type expansions appear in the works of M. Stone and N. Dunford–J. Schwartz in describing resolvents and their poles.

In functional analysis, the theorem acts as an analytic analogue of spectral decomposition in theorems of Gelfand–Mazur, and was used in the classical works of G. H. Hardy on the distribution of zeros of the zeta function. In algebraic geometry, the formulation via sheaf cohomology: the Mittag-Leffler theorem corresponds to the vanishing of the first cohomology of a certain sheaf of meromorphic functions; this is explained in the books by J.-P. Serre and R. Hartshorne. In probability theory and stochastic processes, such meromorphic function representations are applied to characteristic functions and Laplace transforms; examples include works by B. Mandelbrot and Yu. V. Prokhorov on stable distributions, where the analytic continuation is given via poles and residues.

Historical Background and Development of the Idea

The result now known as the Mittag-Leffler theorem was published by Gösta Mittag-Leffler in 1877 in Acta Mathematica and grew out of work on the theory of elliptic functions by Weierstrass and Cauchy. The motivating problem was the construction of "universal" series prescribing the desired arrangement of poles, which allowed systematization of special functions such as elliptic and gamma functions.

In the late 19th century, J. Hadamard and E. Picard used Mittag-Leffler's ideas to study entire functions of finite order of growth; this led to Hadamard's classical factorization theorems. In the early 20th century, after the appearance of Riemann and his school's work on Riemann surfaces, the theorem was reformulated in a geometric spirit; major contributions were made by F. Klein and P. Kusakawa, linking it to the Riemann–Roch theorem.

In the second half of the 20th century, in the works of J.-P. Serre, L. Hörmander, and L. Carleson, Mittag-Leffler's ideas became part of the general language of complex analytic spaces and sheaf cohomology. Proofs began relying on $L^2$-estimate methods and the Dolbeault theorem. In the 21st century, continuations of this line can be seen in problems about longitudinal and transverse meromorphic objects on complex manifolds, as well as in analytic aspects of dynamical systems theory (for example, in studies of meromorphic self-similar functions in the works of C. McMullen and D. Sullivan).