Pell's Equation
Pell's equation is the deceptively simple Diophantine equation
x^2 - D\,y^2 = 1,
for a non-square positive integer D, sought in whole numbers
x, y. It looks like it might have a handful of small solutions — but it
has infinitely many, and some are shockingly large. It is one of the oldest equations in
mathematics, studied by Brahmagupta and Fermat long before John Pell (who, by a historical
mislabelling, lent it his name).
One solution gives them all
There is always a smallest nontrivial solution, the fundamental solution
(x_1, y_1). From it, every other solution is generated by taking powers in
the ring \mathbb{Z}[\sqrt D]:
x_k + y_k\sqrt{D} = \left(x_1 + y_1\sqrt{D}\right)^{k}.
For D = 2, the fundamental solution is
(3, 2) (since 3^2 - 2\cdot 2^2 = 1); squaring
3 + 2\sqrt 2 gives the next solution (17, 12),
and so on forever.
Finding the first solution
How do you find (x_1, y_1)? Through the periodic
continued fraction
of \sqrt{D} — its convergents deliver the fundamental solution directly.
This is essential because brute force is hopeless: for D = 61 the smallest
solution is already
x_1 = 1{,}766{,}319{,}049, \qquad y_1 = 226{,}153{,}980.
Fermat posed exactly this case as a challenge — knowing full well no one would stumble on it by
searching.
Why it matters
Pell's equation is the prototype for understanding units in
rings of algebraic integers:
its solutions are the units of \mathbb{Z}[\sqrt D], and the fact
that they form an infinite cyclic family is a special case of Dirichlet's deep unit theorem. A toy
equation that opens onto the structure of number fields.