Fermat's theorem on sums of two squares

Fermat's theorem on sums of two squares
Name	Fermat's theorem on sums of two squares
Field	Number theory
Statement	Every prime of the form 4n+1 is expressible as a sum of two integer squares
Proved	1640s (claimed)
Prover	Pierre de Fermat (claimed), first complete proof by Leonhard Euler

Fermat's theorem on sums of two squares Fermat's theorem on sums of two squares is a classical result in Number theory asserting that every prime p with p ≡ 1 (mod 4) can be written as p = a^2 + b^2 with integers a and b, while primes p ≡ 3 (mod 4) cannot. The theorem lies at the intersection of the arithmetic of Pierre de Fermat, the work of Leonhard Euler, and later algebraic frameworks from Carl Friedrich Gauss and William Rowan Hamilton. It motivates developments involving Gaussian integer factorization, quadratic reciprocity, and representations by quadratic forms.

Statement of the theorem

Fermat's statement: for any prime p with p ≡ 1 (mod 4) there exist integers a and b such that p = a^2 + b^2; conversely, if p ≡ 3 (mod 4) then p cannot be expressed as a sum of two squares. This links primes in arithmetic progressions studied by Dirichlet's theorem on arithmetic progressions and reciprocity laws exemplified in Quadratic reciprocity law as proved by Carl Friedrich Gauss in his Disquisitiones Arithmeticae.

Historical background and Fermat's proof outline

Pierre de Fermat announced the theorem in correspondence with Blaise Pascal and other contemporaries, claiming a proof that he did not publish. The earliest published complete proof was given by Leonhard Euler, who adapted Fermat's ideas and used congruences and descent methods related to work of Marin Mersenne and Fermat's little theorem. Later expositions by Adrien-Marie Legendre and formalizations by Carl Friedrich Gauss placed the theorem in the context of quadratic forms, while nineteenth-century algebraic number theory from Ernst Kummer and Richard Dedekind reframed it via unique factorization in ring of integers of quadratic fields and Gaussian integers.

Number-theoretic proofs (using Gaussian integers)

A standard proof uses the ring Z[i] of Gaussian integers introduced by Carl Friedrich Gauss and further developed by William Rowan Hamilton conceptually for complex numbers. In Z[i], primes p ≡ 1 (mod 4) split as p = π·π̄ into nonassociate Gaussian primes π, π̄, enabling p = N(π) = a^2 + b^2 where N denotes the norm. This argument depends on unique factorization in Z[i], an instance of the domain concept later axiomatized by Richard Dedekind and Ernst Kummer. Connections to Fermat's little theorem and constructions used by Évariste Galois inform alternatives: exploiting a solution x with x^2 ≡ −1 (mod p) (guaranteed by Quadratic reciprocity law), lifting via the Hensel lemma-style descent yields integers a and b with p = a^2 + b^2. Euler's original descent proof and later treatments by Adrien-Marie Legendre illustrate elementary congruential strategies parallel to the algebraic approach of David Hilbert and Emil Artin.

Consequences and corollaries

The theorem implies characterization of integers representable as sums of two squares: an integer n ≥ 1 is a sum of two squares iff in its prime factorization every prime q ≡ 3 (mod 4) occurs with even exponent, a result tied to Jacobi's two-square theorem and refinements by Srinivasa Ramanujan. It informs the structure of class numbers for imaginary quadratic fields like Q(i) studied by Heegner and Stark; unique factorization in Z[i] makes Q(i) a principal ideal domain, a notion advanced by Claude Chevalley and Emmy Noether. The theorem interacts with results on representation by quadratic forms in Carl Gustav Jacobi's theory and modular form interpretations later explored by Erich Hecke and Atkin.

Examples and computations

Concrete instances: 5 = 1^2 + 2^2 (prime 5 ≡ 1 (mod 4)), 13 = 2^2 + 3^2 (prime 13 ≡ 1 (mod 4)), 17 = 1^2 + 4^2 (prime 17 ≡ 1 (mod 4)), while 3 and 7 (both ≡ 3 (mod 4)) are not sums of two squares. Computational algorithms to find representations use modular square-root computations related to Tonelli–Shanks algorithm and factorization in Z[i] akin to methods in Quadratic sieve and Lenstra elliptic-curve factorization frameworks. Historical tabulations by Adrien-Marie Legendre and later computational verifications by twentieth-century mathematicians like John von Neumann and Paul Erdős illustrate practical checks and enumerations.

Generalizations and related results

Generalizations include characterization of integers represented by forms ax^2 + bxy + cy^2 as in Gauss's theory of binary quadratic forms and the four-square theorem of Joseph-Louis Lagrange. Higher-dimensional analogues and sum-of-k-squares problems connect to Waring's problem and results of David Hilbert. The approach via algebraic number theory extends to primes in other quadratic fields, studied in the context of Hilbert class field theory and reciprocity laws of Emil Artin. Modular and automorphic perspectives relate the theorem to theta-series and investigations by Srinivasa Ramanujan, Atle Selberg, and Yakovlevich.

Category:Theorems in number theory