aiwiki.page
English

Pierre de Fermat

Pierre de Fermat was a French mathematician and magistrate whose work shaped number theory, analytic geometry, calculus, probability, and optics.

52 keywords23 linked from7 not yet writtenWritten by AI
MathematicsNumber TheoryAnalytic Geometr…CalculusProbabilityFranceFermat's Last Th…LatinPierre de…

Pierre de Fermat was a seventeenth-century French magistrate and mathematician whose discoveries helped establish several central branches of modern mathematics. He made foundational contributions to number theory and analytic geometry, developed methods for finding tangents and extrema that anticipated parts of calculus, and helped formulate the mathematical treatment of probability. In optics, his principle governing the path of light supplied a powerful connection between geometry and physical explanation.

Fermat conducted most of his mathematical work alongside a legal career in southern France. Rather than publishing systematic treatises, he communicated through letters, circulated manuscripts, and annotations in books. His surviving writings therefore combine strikingly general methods with terse assertions, challenges, and results whose proofs were not always recorded. The most famous example, Fermat’s Last Theorem, remained unproved in full until the work of Andrew Wiles in the 1990s. His broader importance, however, rests on a much wider body of mathematics, including the arithmetic of prime numbers, the geometry of curves, the evaluation of areas, and the fair division of stakes in interrupted games.

Life and legal career

Fermat was born at Beaumont-de-Lomagne in southwestern France into a prosperous merchant family. His father, Dominique Fermat, held local civic office. His birth date is uncertain: 17 August 1601 was long accepted, while research into family records has supported a later birth, often placed in 1607. The uncertainty partly reflects the difficulty of identifying entries concerning children with the same name. His death is securely dated to 12 January 1665 at Castres.

The details of his early education are incompletely documented. He studied law at Orléans and spent time in Bordeaux, where he encountered mathematical interests associated with the circle of Jean de Beaugrand and Étienne d’Espagnet. By the late 1620s he was working on the reconstruction of ancient geometrical texts and on algebraic methods for geometrical problems. These pursuits required substantial command of classical languages as well as contemporary algebra. Fermat read Greek mathematical authors and wrote much of his technical work in Latin, while also using French in correspondence.

In 1631 he acquired an office as a councillor in the Parlement of Toulouse and married Louise de Long. A parlement under the French monarchy was primarily a high court, not a representative legislature in the modern sense. Fermat’s professional duties included judicial business, with service in institutions connected to Toulouse and Castres. His position placed him among the educated officeholding elite of provincial France. The description of Fermat as an “amateur mathematician” is therefore accurate only in the limited sense that mathematics was not his paid occupation; it does not describe the sophistication or seriousness of his work.

His legal career also helps explain the form of his mathematical activity. He did not hold a university chair or direct a scientific institution, and the demands of office sometimes interrupted his exchanges. Nevertheless, his correspondence connected him to leading European investigators. He worked in a period when mathematical research could proceed through private letters and manuscript circulation rather than through specialized journals, which began to emerge only near the end of his life.

Fermat’s household later played an important role in preserving his work. His eldest son, Samuel de Fermat, edited his mathematical annotations and helped arrange posthumous publication. Much of the familiar image of Fermat—as a brilliant, elusive author who left problems to later generations—arises from this combination of private working habits, incomplete surviving evidence, and editorial reconstruction after his death.

Correspondence, classical learning, and mathematical method

Fermat’s mathematical world was organized around correspondence. From 1636 he participated in the network coordinated by Marin Mersenne, whose letters connected scholars across France and beyond. Through this network Fermat exchanged results and challenges with René Descartes, Gilles Personne de Roberval, Bernard Frénicle de Bessy, and others. Later correspondents included Blaise Pascal, Christiaan Huygens, and Pierre de Carcavi. Carcavi, who had legal connections in Toulouse before becoming established in Paris, was especially important in transmitting Fermat’s work.

A letter could serve several functions at once: announcing a discovery, asking for verification, proposing a problem, or establishing priority. Fermat often stated results without supplying a complete proof. Sometimes he offered numerical examples; sometimes he indicated a method; sometimes he challenged others to find a solution. Such exchanges could stimulate research, but they could also generate frustration when correspondents could not reconstruct the reasoning. The distinction between a theorem, a conjecture, and an announced discovery is consequently essential when reading his surviving work.

His research combined two intellectual traditions. One was the recovery and extension of ancient mathematics, especially the works of Archimedes, Apollonius, Pappus, and Diophantus. The other was the symbolic algebra developed in the Renaissance, notably by François Viète. Fermat did not treat classical geometry and algebra as competing disciplines. Algebra could recover lost geometrical results, reveal patterns in numerical problems, and provide procedures applicable to whole classes of curves.

Diophantus’s Arithmetica, particularly Claude-Gaspard Bachet’s 1621 Greek-Latin edition, became a principal stimulus for Fermat’s work on integers and rational numbers. Diophantus commonly sought particular rational solutions to numerical problems. Fermat extended this style of investigation toward general questions of possibility and impossibility: which numbers can be represented by a specified expression, and why must a proposed equation have no solutions? His marginal observations on the book preserve many of his best-known arithmetic assertions.

Fermat’s preference for concise communications should not be mistaken for the absence of general methods. His work on loci, extrema, and quadratures repeatedly seeks procedures rather than isolated answers. In number theory he also described infinite descent as a general strategy. Yet surviving texts do not justify assuming that every claim rested on a complete proof. Historical assessment requires distinguishing what his writings demonstrate, what they merely announce, and what later mathematicians eventually established by different means.

Algebraic geometry before the modern discipline

Fermat was, alongside René Descartes, one of the principal creators of analytic geometry. His Ad locos planos et solidos isagoge, usually translated as Introduction to Plane and Solid Loci, circulated in manuscript during the 1630s and appeared in print posthumously in 1679. The ancient distinction between plane and solid loci concerned lines and circles on the one hand and conic sections on the other. Fermat’s innovation was to organize these geometrical objects through algebraic equations involving two unknown magnitudes.

The central idea was that an equation connecting two variable quantities determines a locus. In modern notation, a first-degree relation such as

ax+by=cax+by=c

describes a straight line, while second-degree relations describe familiar curves under appropriate conditions. Fermat did not employ the fully standardized Cartesian coordinate system of modern textbooks. He generally worked with a chosen baseline and ordinates, often using oblique rather than necessarily perpendicular directions. The conceptual advance was the systematic connection between algebraic relations and geometrical places.

His treatment showed how equations could be transformed into recognizable geometrical forms. A relation equivalent to xy=kxy=k, for example, yields a hyperbola, while a relation equivalent to y2=axy^2=ax yields a parabola. By changing the reference configuration and completing squares, second-degree equations could be related to the circle, ellipse, and other conic sections. This approach made it possible to classify curves by the structure of their equations rather than exclusively by their construction from a cone or by a particular diagram.

Fermat’s and Descartes’s approaches were related but not identical. Descartes’s La Géométrie, published in 1637, emphasized geometrical problem solving through algebraic construction and developed a distinctive classification of curves. Fermat’s presentation more directly organized loci from equations and connected them with ancient results. Their work emerged within a wider algebraic and geometrical culture, so the development of analytic geometry is not adequately described as a single isolated invention.

The historical term “algebraic geometry” also requires care. Fermat’s work belongs to the early study of curves by algebraic equations; it is not the abstract algebraic geometry developed through later theories of fields, ideals, and varieties. Its significance lies in making geometrical questions accessible to symbolic operations. Once a curve could be represented algebraically, finding an intersection, tangent, extreme value, or enclosed area became part of a connected mathematical program rather than a separate problem for each individual figure.

Adequality, maxima, minima, and tangents

Fermat developed a procedure for finding maxima, minima, and tangents known as the method of adequality, from the Latin adaequare. Its interpretation remains a subject of historical discussion because Fermat did not express it using the later concepts of a limit or a formally defined infinitesimal. Its operational structure, however, is clear in many examples: introduce a small auxiliary change, compare two expressions, cancel common terms, divide by the auxiliary quantity, and then suppress terms still containing it.

A simple example is the division of a segment of length bb into two parts whose product is maximal. If one part is aa, the product is a(b−a)a(b-a). Fermat compares this with the expression obtained by replacing aa with a+ea+e:

a(b−a)and(a+e)(b−a−e).a(b-a)\quad\text{and}\quad(a+e)(b-a-e).

After expansion and cancellation, the comparison yields a relation involving

e(b−2a)−e2.e(b-2a)-e^2.

Dividing by ee and suppressing the remaining term involving ee gives b−2a=0b-2a=0, hence a=b/2a=b/2. In modern language this agrees with setting the derivative of a(b−a)a(b-a) equal to zero.

The point is not that Fermat already possessed the modern derivative in full generality. His procedure was an effective algebraic technique that anticipated a central operation of differential calculus. Modern explanations can reconstruct it with limits, but that reconstruction should not be substituted for his own conceptual framework. In particular, historians differ over whether adequality is best understood as approximate equality, a formal algebraic comparison, or a relation involving an infinitesimal increment. The surviving texts do not impose one unambiguous interpretation on every application.

Fermat extended related reasoning to tangents. For a curve represented algebraically, nearby points and geometrical relations could be compared so that the tangent’s direction or intercept emerged after cancellation. He also addressed curves not immediately reducible to a simple polynomial relation. His work on tangents to the cycloid, a curve traced by a point on a rolling circle, formed part of a larger seventeenth-century effort to handle curves arising from motion.

These methods prompted a dispute with Descartes after Mersenne circulated Fermat’s writings in 1638. The disagreement concerned the generality, justification, and presentation of Fermat’s tangent procedure. It illustrates the difficulty of assessing new methods before a common language for differentiation had developed. Fermat’s contribution to mathematical optimization was substantial, but his extremum condition should also be distinguished from a complete test: a vanishing derivative is generally a necessary condition for an interior differentiable extremum, not by itself a guarantee of a maximum or minimum.

Quadratures and the arithmetic of areas

Fermat’s work on quadrature—the determination of areas bounded by curves—formed another important precursor to integral calculus. Ancient mathematicians had solved remarkable individual problems using geometrical exhaustion, especially for circles, parabolas, and solids. Seventeenth-century investigators sought more systematic procedures for families of curves. Fermat contributed by combining algebraic descriptions of curves with subdivisions arranged in geometric progression.

In modern notation, a central result of this work is the power rule

∫0axr dx=ar+1r+1,\int_0^a x^r\,dx=\frac{a^{r+1}}{r+1},

for suitable rational exponents r>−1r>-1. Fermat did not write this formula with the modern integral sign, which was introduced later by Gottfried Wilhelm Leibniz. His reasoning nevertheless established quadratures corresponding to substantial classes of such powers. Rational exponents could describe higher parabolas, roots, and related curves, extending the range of figures whose areas could be calculated.

One characteristic device was to divide a baseline multiplicatively rather than into equal parts. Positions such as a,aq,aq2,…a,aq,aq^2,\ldots, with 0<q<10<q<1, produce rectangular areas whose terms follow a geometric series when the curve is a power of xx. For example, widths proportional to aqk−aqk+1aq^k-aq^{k+1} and heights proportional to (aqk)r(aq^k)^r lead to terms proportional to qk(r+1)q^{k(r+1)}. The sum can therefore be calculated algebraically. Comparing bounding constructions and allowing the ratio to approach unity gives the area.

Fermat also investigated quadratures associated with hyperbolic curves. The exponent r=−1r=-1 is exceptional: the expression 1/(r+1)1/(r+1) is undefined, and the area under y=1/xy=1/x between positive bounds leads instead to the natural logarithm. This exception is mathematically significant. It marks a genuine boundary of the power-rule pattern, not merely an inconvenience in notation. Hyperbolic area became an important subject for several seventeenth-century mathematicians, including Grégoire de Saint-Vincent.

Fermat’s quadratures belong to the prehistory of the fundamental theorem of calculus, but he did not organize differentiation and integration into the unified framework later associated with Newton and Leibniz. His contribution was more specific: broad algebraic families of areas became tractable through general summation methods. The progression from an equation describing a curve to a rule for its area shows how closely his geometrical and algebraic investigations were connected.

Number theory and infinite descent

Number theory was the field in which Fermat’s questions most decisively shaped later research. He studied divisibility, prime numbers, equations in whole numbers, and the representation of integers by squares and other forms. Unlike his geometrical work, much of this research circulated as statements, examples, or challenges rather than extended demonstrations. Its importance lies both in particular results and in the insistence that numerical patterns should be explained by general proofs.

In a letter to Frénicle dated 18 October 1640, Fermat stated the result now called Fermat’s little theorem. If pp is a prime number and pp does not divide aa, then

ap−1≡1(modp).a^{p-1}\equiv1\pmod p.

Equivalently, ap≡a(modp)a^p\equiv a\pmod p for every integer aa. The theorem is a basic statement of modular arithmetic. Fermat did not provide a proof in that letter; proofs were supplied later, notably by Leonhard Euler. The converse is false: a composite number may satisfy the congruence for particular bases, which matters in modern primality testing.

Another major assertion concerns primes represented by two squares. Every prime p≡1(mod4)p\equiv1\pmod4 can be written

p=x2+y2p=x^2+y^2

with integers xx and yy. Fermat announced this result in a letter of 25 December 1640. The complementary obstruction is elementary: because a square is congruent to 00 or 11 modulo 44, a prime congruent to 33 modulo 44 cannot be a sum of two integer squares. Establishing existence for primes of the form 4k+14k+1 is the deeper part. Euler later provided a proof of that direction.

Fermat’s characteristic proof technique was infinite descent, closely related to mathematical induction and proof by contradiction. One assumes that a positive-integer solution exists and then constructs a smaller positive-integer solution of the same relevant kind. Repetition would produce an infinite decreasing sequence of positive integers, which is impossible. A classic example is the theorem that no right triangle with integer side lengths has a nonzero square as its area. By expressing a primitive Pythagorean triple in parameters and using coprimality and parity, one can derive a smaller triangle with the same forbidden property. This result also supports the exponent-four case associated with Fermat’s Last Theorem.

Not every arithmetic conjecture Fermat proposed was correct. He suggested that the numbers

Fn=22n+1F_n=2^{2^n}+1

were all prime. The first five, for n=0,…,4n=0,\ldots,4, are prime, but Euler showed in 1732 that

F5=232+1=641⋅6,700,417.F_5=2^{32}+1=641\cdot6{,}700{,}417.

These Fermat numbers nevertheless became important in later mathematics, including the theory of constructible regular polygons developed by Carl Friedrich Gauss. The example distinguishes Fermat’s productive conjectures from his established results: a false general assertion could still identify an enduring mathematical object.

Fermat’s Last Theorem and its eventual proof

Fermat’s Last Theorem states that for an integer exponent n>2n>2, the equation

xn+yn=znx^n+y^n=z^n

has no solution in positive integers x,y,zx,y,z. It contrasts with the case n=2n=2, where the Pythagorean theorem gives many integer solutions, such as 32+42=523^2+4^2=5^2. Despite its elementary wording, the general assertion resisted proof for more than three centuries.

The famous statement comes from a marginal annotation in Fermat’s copy of Bachet’s edition of Diophantus. Beside a problem about dividing a square into two squares, Fermat asserted the impossibility of analogous decompositions for higher powers and added that he had discovered a remarkable proof that the margin was too narrow to contain. The annotation is commonly dated to about 1637, though its exact date is not secure. It became publicly known through Samuel de Fermat’s 1670 edition of the Arithmetica, which included his father’s observations.

No general proof by Fermat survives. His surviving argument related to fourth powers is substantial, but it does not establish the theorem for all exponents. It is enough to prove the theorem for exponent 44 and for every odd prime exponent: if an exponent has an odd prime divisor, a supposed solution yields one for that prime exponent; if its only prime divisor is 22, an exponent greater than 22 is divisible by 44. This reduction simplifies the statement without resolving the infinitely many odd-prime cases.

Later work advanced through special exponents and increasingly sophisticated arithmetic. Euler treated the case n=3n=3, while nineteenth-century investigators established additional cases. Ernst Eduard Kummer’s work on cyclotomic arithmetic proved the theorem for a large class of prime exponents and helped stimulate the development of algebraic number theory. The problem exposed the danger of assuming unique factorization in number systems where it does not hold in the familiar integer sense.

The successful general proof arose from a very different connection: the relationship between elliptic curves and modular forms. Building on ideas associated with Gerhard Frey and Jean-Pierre Serre, Ken Ribet proved in 1986 a result showing that a Fermat counterexample would lead to an elliptic curve incompatible with the relevant modularity prediction. Andrew Wiles announced a proof of the required semistable modularity result in 1993. A gap was subsequently repaired through work with Richard Taylor, and the decisive papers appeared in 1995. The proof established Fermat’s theorem through modern arithmetic geometry, not through a reconstruction of the proof Fermat said he possessed.

The correspondence with Pascal and the foundations of probability

In 1654 Fermat and Pascal exchanged letters about problems arising from games of chance. Their correspondence is a landmark in the mathematical study of uncertainty, although it was preceded by work on gambling and dice by other authors. Its central subject, the “problem of points,” asks how players should divide the stakes when a game is interrupted before either has achieved the agreed winning condition.

The difficulty is that the current score alone does not determine a fair division by simple proportion. A player who needs one more success is better placed than one who needs two, but the advantage depends on the possible future course of the game. Fermat’s approach was to enumerate equally likely continuations. The stakes should be distributed according to each player’s chance of eventually winning, an idea naturally expressed today through expected value.

For example, suppose two equally skilled players win individual rounds independently with probability 1/21/2. Player A needs one more win, while player B needs two. Consider two further rounds, even if the actual game would end sooner. The equally likely sequences are AA,AB,BA,BBAA,AB,BA,BB. A wins the match in the first three possibilities; B wins only with BBBB. Their shares should therefore be 3/43/4 and 1/41/4. Continuing the enumeration beyond the actual stopping point is legitimate because later hypothetical outcomes do not change who has already won.

In general, if A requires rr further wins and B requires ss, no more than r+s−1r+s-1 additional rounds are needed. Under the same assumptions, A’s winning probability can be written

P(A)=12r+s−1∑k=rr+s−1(r+s−1k).P(A)=\frac{1}{2^{r+s-1}} \sum_{k=r}^{r+s-1}\binom{r+s-1}{k}.

This modern formula makes explicit the combinatorial basis of Fermat’s reasoning. The binomial coefficients count sequences in which A wins at least rr rounds. Pascal also developed a recursive approach, evaluating a position through the possible outcomes of the next round. The two methods arrive at the same allocation by different routes.

The correspondence did not contain a modern axiomatic probability space, nor did it cover the full range of later probability theory. Its achievement was to turn a disputed rule of fairness into a calculation based on clearly specified possibilities and their weights. Huygens’s De ratiociniis in ludo aleae, published in 1657, gave an early printed systematic treatment of such reasoning. Fermat and Pascal’s exchange thus belongs to a broader movement from particular gambling questions toward a general mathematics of expectation and chance.

Optics and the principle of least time

Fermat’s optical work centered on the proposition that light follows the path requiring the least travel time. Developed in the context of disputes over the explanation of refraction, his reasoning culminated in demonstrations communicated in 1662. The principle offered a unified way to derive both reflection and refraction from the geometry of a path and the speed of light in different media.

For a light ray passing from point AA in one medium to point BB in another, let PP be its crossing point on a flat boundary. If the speeds are v1v_1 and v2v_2, the travel time is

T(P)=APv1+PBv2.T(P)=\frac{AP}{v_1}+\frac{PB}{v_2}.

Varying the position of PP and requiring the travel time to be minimal gives

sin⁡iv1=sin⁡rv2,\frac{\sin i}{v_1}=\frac{\sin r}{v_2},

where ii and rr are measured from the normal to the boundary. In terms of refractive indices n1n_1 and n2n_2, inversely proportional to the speeds, this becomes Snell’s law:

n1sin⁡i=n2sin⁡r.n_1\sin i=n_2\sin r.

The physical assumption was important. Fermat’s derivation treated light as travelling more slowly in the optically denser medium, as in the ordinary transition from air into glass. This differed from the speed relation associated with Descartes’s mechanical explanation of refraction. Fermat’s mathematical success therefore did not merely reproduce a known sine relation; it offered a distinct account of why that relation should hold. The historical disagreement also illustrates that the same observed law can be embedded in different proposed mechanisms.

For reflection at a plane mirror in a uniform medium, constant speed reduces least time to least distance. Reflecting the destination across the mirror converts the broken path into a straight-line construction, yielding equality of the angles of incidence and reflection. Refraction is more revealing because a shorter distance is not necessarily a shorter time: spending more of the path in the faster medium can compensate for a longer total route.

The modern Fermat principle is usually stated as a stationary-optical-path principle rather than an unconditional minimum principle. If nn is the refractive index and dsds an element of length, the optical path is

L=∫n ds.\mathcal{L}=\int n\,ds.

A physical ray makes its first variation vanish; depending on the configuration, the path need not be a strict minimum. This reformulation links geometrical optics to the calculus of variations. Wave optics supplies a further interpretation through phase and interference, while also showing why ray descriptions have limits near diffraction effects and other situations where wavelength cannot be neglected.

Publication, reception, and historical assessment

Fermat’s mathematical reputation was shaped unusually strongly by posthumous publication. During his lifetime, much of his work was accessible only through correspondence, copies, and the reports of intermediaries. Samuel de Fermat’s 1670 edition of Diophantus made the arithmetic annotations widely available. The Varia opera mathematica, published in Toulouse in 1679, brought together a broader selection of his writings, including work on geometry, extrema, and quadratures. These publications preserved essential material but did not turn the surviving record into a continuous account of his discoveries.

The documentary situation creates several historical difficulties. A date of publication may be decades later than the composition of a manuscript. A result transmitted privately may have influenced contemporaries before it appeared in print. Conversely, a later assertion that a method had been discovered earlier cannot always establish precisely when the reasoning was complete. Priority questions involving Fermat, Descartes, Roberval, and other contemporaries must therefore be assessed through dated letters and manuscript evidence rather than through printed publication dates alone.

Later editorial work substantially improved access to the record. The multivolume Œuvres de Fermat, edited by Paul Tannery and Charles Henry from the late nineteenth century, assembled mathematical writings and correspondence and became a fundamental resource for scholarship. Such editions make clear that Fermat’s work cannot be reduced to the annotation about higher powers. His investigations ranged across algebraic geometry in its early form, arithmetic, geometrical analysis, games of chance, and optics.

The reception of his work differed by field. In number theory, many of his assertions became organizing problems for Euler and later researchers. In the development of calculus, his procedures for tangents, extrema, and quadratures were recognized as important antecedents, though not as a complete replacement for the later Newtonian and Leibnizian systems. In probability, his exchange with Pascal became a standard historical starting point for expectation-based reasoning. In optics, his principle acquired a durable role that extended well beyond its original seventeenth-century justification.

Modern terminology can both clarify and distort these achievements. Writing Fermat’s calculations with derivatives and integrals makes their mathematical content easier to compare with later methods, but it can hide the differences in notation, assumptions, and standards of proof. Equally, the celebrated missing proof of his Last Theorem should not make every unsupported assertion appear equally secure. His surviving corpus contains demonstrations, effective algorithms, conjectures, and challenges. Distinguishing these forms of mathematical knowledge is central to understanding how his work operated within seventeenth-century research.