Euler Criterion
Euler criterion gives an efficient way to decide whether an integer is a square modulo an odd prime. Let $p$ be an odd prime and let $a$ be an integer not divisible by $p$....
Detecting Quadratic Residues
Euler criterion gives an efficient way to decide whether an integer is a square modulo an odd prime. Let $p$ be an odd prime and let $a$ be an integer not divisible by $p$. Then $a$ is either a quadratic residue modulo $p$ or a quadratic nonresidue modulo $p$.
The criterion states that this distinction is detected by the power
$$ a^{(p-1)/2}. $$
More precisely,
$$ a^{(p-1)/2}\equiv \begin{cases} 1 \pmod p, & \text{if } a \text{ is a quadratic residue modulo } p,\ -1 \pmod p, & \text{if } a \text{ is a quadratic nonresidue modulo } p. \end{cases} $$
Equivalently, using the Legendre symbol,
$$ a^{(p-1)/2}\equiv \left(\frac{a}{p}\right)\pmod p. $$
Statement of the Theorem
Theorem. Let $p$ be an odd prime and let $a\in\mathbb{Z}$ with $p\nmid a$. Then
$$ a^{(p-1)/2}\equiv \left(\frac{a}{p}\right)\pmod p. $$
This is Euler criterion.
It turns the problem of solving
$$ x^2\equiv a\pmod p $$
into a modular exponentiation problem.
Proof
The nonzero residue classes modulo $p$ form a multiplicative group
$$ \mathbb{F}_p^\times $$
of order
$$ p-1. $$
By Fermat little theorem,
$$ a^{p-1}\equiv1\pmod p. $$
Therefore
$$ \left(a^{(p-1)/2}\right)^2\equiv1\pmod p. $$
So
$$ a^{(p-1)/2}\equiv1\pmod p $$
or
$$ a^{(p-1)/2}\equiv-1\pmod p. $$
Now suppose $a$ is a quadratic residue modulo $p$. Then there exists an integer $x$ such that
$$ a\equiv x^2\pmod p. $$
Hence
$$ a^{(p-1)/2}\equiv (x^2)^{(p-1)/2}=x^{p-1}\equiv1\pmod p. $$
So quadratic residues give the value $1$.
It remains to see that nonresidues give the value $-1$. Since exactly half of the nonzero residue classes modulo $p$ are quadratic residues, there are
$$ \frac{p-1}{2} $$
quadratic residues. The polynomial
$$ X^{(p-1)/2}-1 $$
has degree $(p-1)/2$, and we have already found $(p-1)/2$ roots, namely the nonzero quadratic residues. Therefore these are all its roots in $\mathbb{F}_p$.
Thus any nonresidue cannot satisfy
$$ a^{(p-1)/2}\equiv1\pmod p. $$
Since the only possible values are $1$ and $-1$, every quadratic nonresidue satisfies
$$ a^{(p-1)/2}\equiv-1\pmod p. $$
This proves the theorem.
Example
Determine whether $5$ is a quadratic residue modulo $11$.
Compute
$$ 5^{(11-1)/2}=5^5. $$
Modulo $11$,
$$ 5^2=25\equiv3, $$
$$ 5^4\equiv3^2=9, $$
so
$$ 5^5\equiv9\cdot5=45\equiv1\pmod{11}. $$
Therefore
$$ \left(\frac{5}{11}\right)=1. $$
Hence $5$ is a quadratic residue modulo $11$. Indeed,
$$ 4^2=16\equiv5\pmod{11}, $$
and also
$$ 7^2=49\equiv5\pmod{11}. $$
Example of a Nonresidue
Determine whether $2$ is a quadratic residue modulo $11$.
Compute
$$ 2^5=32\equiv-1\pmod{11}. $$
Therefore
$$ \left(\frac{2}{11}\right)=-1. $$
So $2$ is a quadratic nonresidue modulo $11$. The congruence
$$ x^2\equiv2\pmod{11} $$
has no solution.
Computational Use
Euler criterion is effective because modular exponentiation can be performed quickly by repeated squaring.
For large primes, checking all squares modulo $p$ would be inefficient. Euler criterion reduces the problem to computing one power modulo $p$.
For example, to test whether $a$ is a quadratic residue modulo a large prime $p$, compute
$$ a^{(p-1)/2}\pmod p. $$
If the result is $1$, then $a$ is a residue. If the result is $p-1$, which represents $-1$, then $a$ is a nonresidue.
This is useful in computational number theory, cryptography, and primality testing.
Relation to Group Theory
Euler criterion reflects the structure of the cyclic group
$$ \mathbb{F}_p^\times. $$
Since this group has order $p-1$, its subgroup of squares has order
$$ \frac{p-1}{2}. $$
The map
$$ x\mapsto x^2 $$
has kernel
$$ {1,-1}, $$
so its image has exactly half of the nonzero elements.
The function
$$ a\mapsto a^{(p-1)/2} $$
is a group homomorphism from
$$ \mathbb{F}_p^\times $$
to
$$ {1,-1}. $$
Its kernel is precisely the subgroup of squares. Thus Euler criterion identifies the quadratic residue character of the finite field.
Conceptual Meaning
Euler criterion translates a solvability question into a power congruence. Instead of asking directly whether
$$ x^2\equiv a\pmod p $$
has a solution, it asks whether $a$ lies in the kernel of the character
$$ a\mapsto a^{(p-1)/2}. $$
This viewpoint prepares the way for quadratic reciprocity, Dirichlet characters, and the use of multiplicative characters in analytic number theory.