PlanetMath (more info)
 Math for the people, by the people. Sponsor PlanetMath
Encyclopedia | Requests | Forums | Docs | Wiki | Random | RSS  
Login
create new user
name:
pass:
forget your password?
Main Menu
Owner confidence rating: High Entry average rating: No information on entry rating
[parent] proof of Euler's criterion (Proof)

(All congruences are modulo $p$ for the proof; omitted for clarity.)

Let$$ x = a^{(p-1)/2}$$

Then $x^2 \equiv 1$ by Fermat's Little Theorem. Thus:$$ x \equiv \pm 1$$

Now consider the two possibilities:

  • If $a$ is a quadratic residue then by definition, $a \equiv b^2$ for some $b$ . Hence:$$ x \equiv a^{(p-1)/2} \equiv b^{p-1} \equiv 1$$
  • It remains to show that $a^{(p-1)/2} \equiv -1$ if $a$ is a quadratic non-residue. We can proceed in two ways:
    Proof (a)
    Partition the set $\setof{1, \ldots, p - 1}$ into pairs $\setof{c, d}$ such that$c d \equiv a$ . Then $c$ and $d$ must always be distinct since $a$ is a non-residue. Hence, the product of the union of the partitions is:$$ (p - 1)! \equiv a^{(p-1)/2} \equiv -1$$

    and the result follows by Wilson's Theorem.

    Proof (b)
    The equation:$$ z^{(p-1)/2} \equiv 1$$

    has at most $(p-1)/2$ roots. But we already know of $(p-1)/2$ distinct roots of the above equation, these being the quadratic residues modulo $p$ . So $a$ can't be a root, yet $a^{(p-1)/2} \equiv \pm 1$ . Thus we must have:$$ a^{(p-1)/2} \equiv -1$$

QED.




"proof of Euler's criterion" is owned by Koro. [ full author list (2) | owner history (2) ]
(view preamble | get metadata)

View style:


This object's parent.
Log in to rate this entry.
(view current ratings)

Cross-references: QED, roots, equation, Wilson's theorem, union, product, partition, quadratic non-residue, quadratic residue, Fermat's little theorem, proof, congruences

This is version 4 of proof of Euler's criterion, born on 2002-05-31, modified 2004-03-20.
Object id is 2980, canonical name is ProofOfEulersCriterion.
Accessed 5775 times total.

Classification:
AMS MSC11A15 (Number theory :: Elementary number theory :: Power residues, reciprocity)

Pending Errata and Addenda
None.
[ View all 3 ]
Discussion
Style: Expand: Order:
forum policy
I am a sad panda. by liyang on 2009-04-21 02:21:31
Bloody hell. I wrote this? I wish I still understood it. I wish I had the spare time to understand it again.
[ reply | up ]
an exposition on qr by drini on 2002-06-02 02:06:12
A brief talk on quadratic residues and quadratic reciprocity, including this criterion (with proof)
it's found at Expositions section.

However it's in spanish. The entry is called "Reciporicidad Cuadrática".
 f
G -----> H G
p \ /_ ----- ~ f(G)
 \ / f ker f
 G/ker f 
[ reply | up ]

Interact
post | correct | update request | add example | add (any)