Regular Prime

In the vast landscape of mathematics, some concepts act as keys, unlocking hidden passages between seemingly unrelated worlds. The theory of ​​regular primes​​ is one such master key. Born from a "beautiful mistake" in an attempt to solve the centuries-old puzzle of Fermat's Last Theorem, the idea of regularity evolved from a clever patch into a foundational concept in modern number theory. It addresses a fundamental roadblock in mathematics: the failure of unique factorization in more complex number systems. This article demystifies regular primes, tracing their journey from a historical curiosity to a central player in some of the most profound mathematical theories of our time.

This exploration is structured to build your understanding layer by layer. First, in the "Principles and Mechanisms" section, we will delve into the heart of the matter. We will uncover what regular primes are, how they are defined in terms of class numbers, and how Ernst Kummer's miraculous criterion connects them to the seemingly random Bernoulli numbers. Following that, the "Applications and Interdisciplinary Connections" section will broaden our perspective, showcasing how this single idea served as the crucial tool for proving Fermat's Last Theorem for a large class of primes and built bridges to complex analysis, computational mathematics, and the grand unifying framework of Iwasawa theory.

Alright, let's get our hands dirty. We've been introduced to the idea of ​​regular primes​​, but what are they really? And more importantly, why would anyone care? As with so many great ideas in physics and mathematics, the story begins not with a success, but with a beautiful, spectacular failure. It’s a story about trying to solve an ancient puzzle, stumbling into a hidden world of numbers, and finding a surprising connection that nobody expected.

The puzzle was, of course, Fermat's Last Theorem. You know the one: the equation $x^n + y^n = z^n$ has no whole number solutions for $x, y, z$ (other than the trivial ones) when the exponent $n$ is greater than $2$. For centuries, mathematicians had chipped away at it, proving it for this exponent and that.

Then, in 1847, the French mathematician Gabriel Lamé announced a general proof. His idea was brilliant. He worked not with ordinary integers, but in a richer world of numbers called ​​cyclotomic fields​​. For a prime exponent $p$, he used numbers of the form $a_0 + a_1\zeta_p + a_2\zeta_p^2 + \dots + a_{p-2}\zeta_p^{p-2}$, where the $a_i$ are integers and $\zeta_p$ is a primitive $p$-th root of unity (think of it as a complex number that, when raised to the $p$-th power, gives $1$). In this world, the equation $x^p + y^p = z^p$ can be factored splendidly:

$(x+y)(x+y\zeta_p)(x+y\zeta_p^2)\cdots(x+y\zeta_p^{p-1}) = z^p$

Lamé's plan was simple: if these numbers behave like ordinary integers—specifically, if they have unique factorization into primes—then the factors on the left must be $p$-th powers of other numbers in this world (up to some details). This puts an enormous constraint on $x$ and $y$, so much so that he could show it was impossible. It was an elegant and powerful argument.

There was just one problem. It was wrong.

Another mathematician in the audience, Joseph Liouville, pointed out a gaping hole in the logic. Lamé had assumed that unique factorization, a property we take for granted with whole numbers, also worked in his new world of $\mathbb{Z}[\zeta_p]$. It often doesn't! It’s like discovering that in a foreign country, a dollar bill isn't always worth 100 cents; sometimes it’s worth 98, sometimes 103, and the rules are complicated.

This is where the German mathematician Ernst Eduard Kummer enters the story. He had been working on this very problem for years and understood the subtleties. He knew unique factorization failed. But instead of giving up, he asked a new question. What if we can’t fix the problem for all primes $p$? What if we can fix it for some of them? What if we can identify the "good" primes where the failure of unique factorization is manageable?

This is the birth of the idea of regular primes. Kummer’s genius was to find a workaround, a way to restore just enough of the unique factorization property to make his proof of Fermat's Last Theorem work for a large class of primes.

So, how "badly" does unique factorization fail? Mathematicians have a wonderful tool to measure this. For any given number field (like $\mathbb{Q}(\zeta_p)$), they construct an object called the ​​ideal class group​​, $\mathrm{Cl}(K_p)$. You don't need to know the technical details, but you can think of it as a small, finite group whose very existence is a testament to the failure of unique factorization.

If the class group is trivial (it has only one element), then hurrah! Unique factorization holds. If the class group is not trivial, unique factorization fails. The size of this group, an integer called the ​​class number​​, $h(K_p)$, tells you how messy things are.

Kummer’s key insight was that for his proof of Fermat's Last Theorem, the total failure of unique factorization wasn’t the problem. The real obstacle was when the class number $h(K_p)$ was divisible by the prime exponent $p$. If $p$ does not divide the class number, the specific kind of trouble that would derail his proof is absent. He called such primes ​​regular​​.

So, we have our first real definition: A prime $p$ is ​​regular​​ if and only if it does not divide the class number $h(\mathbb{Q}(\zeta_p))$.

Technically, this means the $p$-primary part of the class group is trivial. In simpler terms, there are no elements in the class group whose order is $p$. This has a crucial consequence: if you have an ideal $\mathfrak{a}$ whose $p$-th power, $\mathfrak{a}^p$, is a principal ideal (the kind of ideal generated by a single number, which acts as our stand-in for a number itself), then the regularity condition forces $\mathfrak{a}$ itself to be a principal ideal. This is Kummer’s powerful substitute for full unique factorization.

This is great, but it seems we've just traded one impossible problem for another. Calculating class numbers is notoriously difficult! How could Kummer possibly check this condition for any given prime $p$?

And now for the magic trick. This is the part of the story that, even today, feels utterly astonishing. Kummer found a criterion for regularity that had nothing to do with class groups or cyclotomic fields. It had to do with a quirky sequence of numbers that appear in a completely different branch of mathematics: calculus.

These are the ​​Bernoulli numbers​​, $B_k$. They were first discovered when trying to find a formula for the sum of powers, $1^k + 2^k + \dots + n^k$. They are defined by a generating function, a kind of power series clothesline on which we hang the numbers: $\frac{t}{\exp(t)-1} = \sum_{n=0}^{\infty} B_n \frac{t^n}{n!}$

Let's not worry about the formula. Let's just look at the first few. Using a recurrence relation derived from this definition, we can compute them: $B_0 = 1, B_1 = -\frac{1}{2}, B_2 = \frac{1}{6}, B_3 = 0, B_4 = -\frac{1}{30}, B_5 = 0, B_6 = \frac{1}{42}, \dots$

They seem like a random jumble of fractions. What on earth could they have to do with prime numbers and unique factorization? Kummer's incredible discovery, now known as ​​Kummer's Criterion​​, is this:

An odd prime $p$ is ​​regular​​ if and only if it does not divide the numerator of any of the Bernoulli numbers $B_2, B_4, \dots, B_{p-3}$.

This is unbelievable! We've replaced a deep, abstract algebraic condition with a finite arithmetic check. To see if a prime $p$ is regular, you just have to compute a handful of these weird fractions and check their numerators. For instance, to check if $p=13$ is regular, we need to inspect the numerators of $B_2, B_4, B_6, B_8, B_{10}$. They are $1, -1, 1, -1, 5$ respectively. None is divisible by $13$, so $13$ is a regular prime.

A prime that is not regular is called ​​irregular​​. The number of Bernoulli numbers in the list whose numerators are divisible by $p$ is called the ​​index of irregularity​​ of $p$, denoted $i(p)$. A prime is regular if and only if its index of irregularity is zero. The first irregular prime is $p=37$, because $37$ divides the numerator of $B_{32}$ (which is a large number!).

Thanks to this criterion, Kummer was able to prove that the first case of Fermat's Last Theorem holds for all regular primes.

Why does this work? The connection is deep and relies on the hidden structure of the class group. The field $\mathbb{Q}(\zeta_p)$ contains complex numbers, so we have a natural operation: complex conjugation, which sends a number $a+bi$ to $a-bi$. In our field, it sends $\zeta_p$ to $\zeta_p^{-1}$.

This operation splits the $p$-part of the class group into two pieces: a "plus" part, $\mathrm{Cl}^+$, which is fixed by conjugation, and a "minus" part, $\mathrm{Cl}^-$, which is inverted by conjugation. The corresponding class numbers are called $h^+$ and $h^-$.

The amazing fact, established by the Herbrand-Ribet theorem, is that the irregularity detected by Kummer's criterion lives entirely inside the 'minus' part. Specifically, an odd prime $p$ is irregular if and only if it divides the relative class number $h^-$. The Bernoulli numbers $B_2, B_4, \dots$ are precisely the tools that probe the structure of this minus part. In fact, it's even more precise: the divisibility of a specific Bernoulli number $B_k$ by $p$ corresponds to a specific "eigenspace" within the minus part of the class group being non-trivial.

And what about $h^+$? This part of the class number is much more mysterious. A famous unsolved problem, ​​Vandiver's Conjecture​​, predicts that $p$ never divides $h^+$. While the connection between $p$ and $h^-$ is beautifully explained by Bernoulli numbers, the structure of $h^+$ is controlled by a different, more subtle set of objects.

The story doesn't end there. The connection between Bernoulli numbers and the class group is just one thread in a much larger tapestry. Those same Bernoulli numbers also appear as special values of the ​​Riemann zeta function​​, the function at the heart of the most famous open problem in mathematics. For an even integer $k \ge 2$, we have the beautiful formula $\zeta(1-k) = -B_k/k$.

This means Kummer's criterion can be rephrased: $p$ is irregular if $\zeta(1-k) \equiv 0 \pmod{p}$ for some even $k$ in the range $2 \le k \le p-3$. Suddenly, our problem about whole numbers is connected to the world of complex analysis.

Modern number theory takes this even further. For each prime $p$, one can construct a ​​$p$-adic L-function​​, $L_p(s, \chi)$, which is like a version of the Riemann zeta function that lives in the world of $p$-adic numbers. These functions have an amazing "interpolation property"—they smoothly connect the classical values related to Bernoulli numbers. The condition that $p$ divides the numerator of $B_k$ is exactly equivalent to a special value of one of these $p$-adic L-functions being divisible by $p$ in the $p$-adic world.

What began as a clever trick to patch a hole in a proof of Fermat's Last Theorem has evolved into a central pillar of modern algebraic number theory, revealing a profound and unexpected unity between integers, complex analysis, and the strange-but-wonderful world of $p$-adic numbers.

And what about $p=2$? The whole theory is for odd primes. It turns out that for $p=2$, the entire structure degenerates into triviality. The "cyclotomic field" $\mathbb{Q}(\zeta_2)$ is just the rational numbers $\mathbb{Q}$, its class number is $1$, the Galois group is trivial, and the list of Bernoulli numbers to check is empty. The problem Kummer was trying to solve simply doesn't exist for $p=2$. The real action, and the real beauty, begins with $p=3$.

Now that we have grappled with the definition of a regular prime—a curious condition tied to the esoteric Bernoulli numbers—you might be wondering, "What is all this for?" It is a fair question. The world of pure mathematics can sometimes feel like a gallery of strange and beautiful sculptures, disconnected from everything else. But the story of regular primes is a spectacular exception. It is a story about how one simple-looking idea, born from an attempt to solve an ancient puzzle, became a Rosetta Stone for translating between wildly different fields of mathematics. It is a journey that will take us from number theory's most famous unsolvable problem to the frontiers of complex analysis, computational science, and the grand, unifying theories of the 21st century. So, let us begin.

For centuries, the brightest minds in mathematics were haunted by a simple statement left in the margins of a book by Pierre de Fermat. He claimed that the equation $x^n + y^n = z^n$ has no solutions in whole numbers for any power $n$ greater than 2. This puzzle, which came to be known as Fermat's Last Theorem, resisted all attacks.

In the 19th century, the German mathematician Ernst Kummer had a revolutionary insight. His idea was to change the rules of the game. Instead of working with ordinary integers, he decided to factor the expression $x^p + y^p$ (for a prime exponent $p$) in a new, richer number system called a cyclotomic field, the field $\mathbb{Q}(\zeta_p)$ generated by a primitive $p$-th root of unity, $\zeta_p$. In this world, the equation splits into a product of $p$ factors: $x^p + y^p = (x+y)(x+\zeta_p y)(x+\zeta_p^2 y)\cdots(x+\zeta_p^{p-1} y)$ Kummer hoped to show that if this product equals a $p$-th power ($z^p$), then each factor must itself be a $p$-th power. This strategy works beautifully for ordinary integers, thanks to the fundamental theorem of arithmetic—unique prime factorization. However, Kummer soon discovered a tragic flaw: in these new number systems, unique factorization can fail!

The culprit behind this failure is a fascinating object we now call the ​​ideal class group​​. You can think of it as a small, finite group that measures exactly how badly unique factorization breaks down. If this group is trivial, all is well. If it is not, the proof stalls.

This is where regular primes make their heroic entrance. A prime $p$ is regular if and only if it does not divide the order of the ideal class group of $\mathbb{Q}(\zeta_p)$. For such primes, the part of the class group that could cause trouble for an exponent of $p$ is trivial. This was exactly the key Kummer needed. He proved that for any regular prime exponent $p$, the first case of Fermat's Last Theorem (where $p$ does not divide $x, y,$ or $z$) holds true. With this single concept, he proved Fermat's Last Theorem for a vast new class of prime exponents, a monumental achievement.

But mathematics is never so simple. Kummer found that not all primes are regular. Is regularity a necessary condition for his proof? The story takes another twist. The first case of Fermat's Last Theorem can still hold even for some irregular primes. For example, the smallest irregular prime is 37, yet the first case of the theorem holds for it. This suggests that the structure of the ideal class group is more subtle than a simple regular/irregular dichotomy. Regularity was not the end of the story, but the beginning of a deeper exploration.

The existence of irregular primes like 37 that still behaved "well" for Fermat's Last Theorem prompted mathematicians to look closer at the class group. The group, it turns out, has an internal structure. It can be split into two pieces: a "plus" part, connected to the maximal real subfield $\mathbb{Q}(\zeta_p + \zeta_p^{-1})$, and a "minus" part.

This led to a famous and long-standing question known as ​​Vandiver's Conjecture​​. The conjecture proposes that for any prime $p$, the "plus" part of the class group, $h(K^+)$, is never divisible by $p$. In other words, all the "irregularity"—all the divisibility by $p$ that messes up our naive assumptions—is locked away in the "minus" part. Kummer had already shown that if Vandiver's Conjecture holds for a prime $p$ (even an irregular one), the first case of Fermat's Last Theorem follows. Since the conjecture has been computationally verified for all primes up to enormous bounds, this provided a much stronger tool than regularity alone. For over a century, this conjecture stood as a tantalizing guidepost, showing how the initial study of regular primes led to more refined questions about the intricate architecture of number fields.

If the story ended there, it would already be a beautiful chapter in the history of mathematics. But the true magic of regular primes lies in their uncanny ability to appear where you least expect them. Let’s leave the world of algebraic number theory for a moment and journey into the realm of analysis, to a celebrity of the mathematical world: the Riemann zeta function, $\zeta(s) = \sum_{n=1}^\infty \frac{1}{n^s}$.

What could this infinite sum possibly have to do with whether a prime is regular? The answer lies in a stunning formula discovered by Leonhard Euler, which connects the values of the zeta function at even integers to none other than our friends, the Bernoulli numbers. A beautiful way to see this connection is by writing down two different expansions for the function $\pi \cot(\pi x)$: one using a sum over integers derived from its poles, and another using its relationship with the exponential function, which brings in the Bernoulli numbers. By comparing the coefficients of these two series, a miraculous formula emerges: $\zeta(2n) = (-1)^{n+1} \frac{(2\pi)^{2n} B_{2n}}{2(2n)!}$ This formula is a bridge between two worlds. On one side, we have analysis ($\zeta(s)$ and $\pi$). On the other, we have number theory and combinatorics ($B_{2n}$).

Let's see this bridge in action. We know that the value of $\zeta(12)$ involves the number 691. Using the formula above for $n=6$, we can work backward and calculate the 12th Bernoulli number, $B_{12}$. When you do the arithmetic and cancel all the common factors, you find that $B_{12} = - \frac{691}{2730}$. There it is! The prime 691 appears in the numerator. And what does Kummer's criterion tell us? It says that a prime $p$ is irregular if it divides the numerator of some $B_k$ for $k \le p-3$. Since $12 \le 691-3$, the prime 691 is indeed an irregular prime! An innocent-looking value of the zeta function contains a deep secret about the arithmetic of the 691st cyclotomic field. This is the kind of profound, unexpected unity that makes mathematics so breathtaking.

So far, we have talked about discovering regular and irregular primes as if it were a matter of course. But how does one actually do it? How did Kummer find that 37 is the first irregular prime? You can't just stare at the class group.

This is where the connection to Bernoulli numbers becomes a powerful, practical tool. The definition of Bernoulli numbers via their generating function leads to a recurrence relation that allows us to compute them one by one. To test if a prime $p$ is regular, we don't need the full Bernoulli numbers, which have enormous numerators and denominators. We only need to know if their numerators are divisible by $p$. This means we can do all the calculations modulo $p$.

The algorithm is then straightforward: for a given prime $p$, we compute the values $B_2 \pmod p, B_4 \pmod p, \dots, B_{p-3} \pmod p$. If any of these values are zero, the prime is irregular. If they are all non-zero, the prime is regular. For instance, a direct computation shows that for $p=23$, none of the relevant Bernoulli numbers are divisible by 23, so 23 is regular. When we run this same process for $p=37$, we find that $B_{32}$ is divisible by 37, revealing its irregularity.

What was once a monumental task for Kummer can now be performed by a computer in a fraction of a second. This interplay between abstract theory and concrete computation is a hallmark of modern number theory. The quest to understand regular primes drives the development of new algorithms, and in turn, computational evidence provides data and intuition for new theoretical conjectures.

The story, remarkably, does not stop here. In the 20th century, the concept of regularity was absorbed into a revolutionary new framework known as ​​Iwasawa theory​​. The central idea of Iwasawa theory is to study not just a single cyclotomic field $\mathbb{Q}(\zeta_p)$, but an entire infinite tower of fields: $\mathbb{Q}(\zeta_p), \mathbb{Q}(\zeta_{p^2}), \mathbb{Q}(\zeta_{p^3}), \dots$. The theory seeks to understand how the class groups behave as one climbs this tower.

All the $p$-parts of these class groups are packaged into a single, magnificent object called the ​​Iwasawa module​​, which we can call $X$. It's a kind of "master" class group containing information about the entire tower. The spectacular ​​Main Conjecture of Iwasawa Theory​​, now a celebrated theorem thanks to the work of Barry Mazur and Andrew Wiles, makes an incredible claim: this purely algebraic object ($X$, or at least its minus part) is perfectly described by a purely analytic object called a ​​$p$-adic L-function​​. This L-function is a $p$-adic analogue of the Riemann zeta function, and its properties are governed by generalized Bernoulli numbers.

In this breathtakingly modern language, what does it mean for a prime $p$ to be regular? It means that the class group at the very bottom of the tower is trivial (its $p$-part, at least). The machinery of Iwasawa theory shows that this is equivalent to the entire minus part of the Iwasawa module, $X^-$, being trivial. And what does the Main Conjecture say about this? It says that $X^-$ is trivial if and only if its corresponding $p$-adic L-function is a "unit"—a non-zero element in a special algebraic sense.

Think about what this means. The simple, classical condition of regularity, which Kummer defined using Bernoulli numbers, corresponds to the most trivial possible case in a vast, modern, and highly complex theory. It’s like discovering that the fundamental note of a simple flute is also the foundational vacuum state in a grand quantum field theory. The concept of a regular prime, once a specialized tool for a single problem, turned out to be a key that unlocked a deep and universal principle governing the arithmetic of number fields.

From a puzzle in a margin, to a flaw in a proof, to a bridge between disciplines, and finally to the cornerstone of a modern mathematical edifice—the journey of regular primes is a testament to the interconnected, ever-expanding beauty of the mathematical universe.