Pi hiding in prime regularities

A hidden arithmetic regularity—how primes split inside the Gaussian integers—turns a messy lattice-point counting problem into a clean alternating series for π. The key move is to count integer grid points inside a large circle in two ways: geometrically, the count is about πr², but algebraically it can be expressed through which integers can be written as a sum of two squares. Once that algebraic count is reorganized using a special multiplicative function tied to primes mod 4, the circle’s area forces π to equal a simple alternating infinite sum.

The setup begins with lattice points (a,b) with integers a and b. Points at distance √n from the origin correspond exactly to solutions of a²+b²=n. Translating (a,b) into the complex number a+bi reframes the condition as (a+bi)(a−bi)=n, so the problem becomes one of factoring n inside the ring of Gaussian integers. In that ring, primes behave in a structured way: primes congruent to 1 mod 4 (like 5, 13, 17) split into two distinct Gaussian primes, while primes congruent to 3 mod 4 (like 3, 7, 11) remain “unsplittable” even after allowing i. This splitting/non-splitting dichotomy decides whether a given n has any representation as a²+b² and, if it does, how many lattice points lie on the circle of radius √n.

From there, the counting becomes a combinatorics-of-factorization recipe. For each prime power dividing n, the number of ways to distribute Gaussian prime factors between two conjugate columns determines the number of lattice points on the circle. A factor of 2 is special but ultimately doesn’t change the count: powers of 2 contribute no net increase because swapping conjugate Gaussian primes only rotates the final complex number by multiples of 90 degrees.

To package the prime-power rules efficiently, the argument introduces a function χ on natural numbers. It is defined so that χ(p)=1 when p≡1 (mod 4), χ(p)=−1 when p≡3 (mod 4), and χ(n)=0 for even n. χ is multiplicative, meaning χ(ab)=χ(a)χ(b), which allows the lattice-point count for radius √n to be written as 4 times a sum of χ over all divisors of n. When lattice points inside a circle of radius r are counted by summing over all n up to r², the divisor-sum structure can be reorganized by columns: for large r, the contribution from each divisor d is roughly proportional to r²/d. This collapses into the alternating series 1−1/3+1/5−1/7+… .

Since the geometric count of lattice points inside the circle is πr² up to smaller error, dividing by r² and letting r grow forces that alternating series to converge to π. The result is a rare bridge between geometry, complex multiplication, and prime factorization—showing that π emerges from the same mod-4 regularity that governs which primes split in the Gaussian integers.