A research on the theory of residue of higher degree
Team members Yanjie Jing Tao Gu
Teacher Hongliang Shi
School
No.2 Secondary School Attached to East China Normal University
Statement
When we participated in the national competition, we were told that all of the results have been discussed in some professional books on number theory, and there have been far better methods to solve them. We didn’t know it when we were doing the study. These results and proofs are all got by ourselves.
Abstract
In this paper, we first discussed some properties of the symbol
p
ka ⎟⎟ ⎠
⎜⎜ ⎞
⎝
⎛
to make the following discussionmore convenient. Then we used the methods of elementary number theory to discuss some basic
properties of residue of higher degree on condition that the degree
k
is an odd prime,2
n,p
n and any positive integer.
Key word: residue of higher degree
Introduction
The conception of quadratic residue comes from the problem of solving quadratic congruence. A general quadratic congruence to a particular
modulo
ax
2+ bx + c ≡ 0 (mod p )
can always be converted into the basic two terms congruencex
2≡ a (mod p )
. As for the solvability of it, Euler once got a effective criterion.To be convenient, Legendre introduced the symbol
⎟⎟
⎠
⎜⎜ ⎞
⎝
⎛ p
a
. So, the problem is to calculate thesymbol
⎟⎟ ⎠
⎜⎜ ⎞
⎝
⎛ p
a
. Early in a former paper, Euler raised a conjecture. Both Legendre and Gaussrealized the importance of the conjecture for calculating the value of Legendre symbol. The
conjecture was finally completely proved by Gauss. That is the famous law of quadratic reciprocity. Thus, the problem of calculating Legendre symbol has been completely solved.
But solving congruence equation of higher degree, even the most basic two terms congruence, is a really difficulty in elementary number theory. In this paper, we discussed some basic properties of two terms congruence equation
x
k≡ a (mod p )
and some related problems with theories of congruence. Although most of these conclusions can be easily got with algebraic number theory, we mainly used elementary methods.
Definition First we suppose that( , )a p =1
. If
x
k≡ a (mod p )
has solutions, we calla
isthe residue of
k
degree modulop
, otherwise we calla
the non‐residue ofk
degreemodulo
p
. For convenience, we introduced the symbolp
ka ⎟⎟
⎠
⎜⎜ ⎞
⎝
⎛
. We denote
= 1
⎟⎟ ⎠
⎜⎜ ⎞
⎝
⎛ p
ka
, it
represents that
a
is the residue modulop
of degreek
; If− 1
⎟⎟ =
⎠
⎜⎜ ⎞
⎝
⎛ p
ka
, it represents that
a
is the non‐residue modulop
of degreek
. Ifk = 2
, then it is quadratic residue,which we are familiar with. And at this time we can omit
k
, and briefly symbol it as⎟⎟ ⎠
⎜⎜ ⎞
⎝
⎛ p a
. If
1
) ,
( a p >
, we denote⎟⎟ ⎠ = 0
⎜⎜ ⎞
⎝
⎛ p
ka
. This condition is special, we didn’t discuss it in this paper.
Main results
In this paper, we always postulate
p
as a prime. The conclusion whenp = 2
is commonly known,(we always have1
2 ⎟ =
⎠
⎜ ⎞
⎝
⎛
k
a
).Thus, we always postulatep
as an odd prime.
1. Properties of the symbol
p
ka ⎟⎟
⎠
⎜⎜ ⎞
⎝
⎛
First, we will discuss some properties of the symbol
p
ka ⎟⎟
⎠
⎜⎜ ⎞
⎝
⎛
to make the following discussionmore convenient.
Obviously, we have
1 ⎟⎟ ⎠ = 1
⎜⎜ ⎞
⎝
⎛
p
k ,⎟⎟ ⎠ = 1
⎜⎜ ⎞
⎝
⎛
k k
p
a
, and ifa ≡ b (mod p )
,k
k
p
b p
a ⎟⎟ ⎠
⎜⎜ ⎞
⎝
= ⎛
⎟⎟ ⎠
⎜⎜ ⎞
⎝
⎛
.It is easy to popularize the following Euler criterion:
Theorem 1 If
p ≡ 1 (mod k )
, then we have(1)
1 1 (mod )
1
p p a
a
pkk
≡
⇔
⎟⎟ =
⎠
⎜⎜ ⎞
⎝
⎛
−; (2)
1 1 (mod )
1
1 1
p p a
a
kn k n p
k
−
≡
⇔
−
⎟⎟ =
⎠
⎜⎜ ⎞
⎝
⎛ ∑
−=
−
.
Proof :If
p ≡ 1 (mod k )
, ifx
k≡ a (mod p )
has solutionx ≡ x
1(mod p )
,then
( ) (mod )
1 1
1
a p
x
kp k p k
−
−
≡
, that is 1 11 (mod )
1
p x
a
k pp
≡
≡
−−
. On the contrary, if
1 (mod )
1
p a
kp
≡
−
, we can know
( ) ( ) ( )
1
1 1 1 1
1 1
( 1) 1 1
1
0 1 1 ( ) (mod )
p
p p p p
p p k i i
p k k k k k k k k k k
i
x x a a x a x a x a p
− − − − −
− − − − −
−
=
≡ − = − + − ≡ − = −
∑
.
Hence there must exist solutions to
x
k− a ≡ 0 (mod p )
. And we can also know that there arek
solutions tox
k− a ≡ 0 (mod p )
.If
p ≡ 1 (mod k )
, we have1 ( 1 ) 0 (mod )
1
0 1 1
1
a a p
a
k
n k n p k
p
p
− = − ∑
−≡
=
−
−
−
If
⎟⎟ ⎠ = − 1
⎜⎜ ⎞
⎝
⎛ p
ka
,1 \ 0 (mod )
1
p a
kp
≡
−
−
, so
1 (mod )
1
1 1
p a
k
n k n p
−
∑
−≡
=
−
.
On the contrary, if
1 (mod )
1
1 1
p a
k
n k n p
−
∑
−≡
=
−
, we can know
1 \ 0 (mod )
1
p a
kp
≡
−
−
, thus
− 1
⎟⎟ =
⎠
⎜⎜ ⎞
⎝
⎛ p
ka
.In quadratic residue there is
⎟⎟
⎠
⎜⎜ ⎞
⎝
= ⎛
⎟⎟ ⎠
⎜⎜ ⎞
⎝
⎟⎟ ⎛
⎠
⎜⎜ ⎞
⎝
⎛
p ab p
b p
a
, this is not surely tenable in residue ofhigher degree, next we will proof :
Theorem 2
∀ a, b
, ifk
k
p
b p
a ⎟⎟
⎠
⎜⎜ ⎞
⎝
⎟⎟ ⎛
⎠
⎜⎜ ⎞
⎝
⎛ ,
don’t both equal− 1
,k k
k
p
ab p
b p
a ⎟⎟
⎠
⎜⎜ ⎞
⎝
= ⎛
⎟⎟ ⎠
⎜⎜ ⎞
⎝
⎟⎟ ⎛
⎠
⎜⎜ ⎞
⎝
⎛
is alwaystenable.
Proof :If
⎟⎟ ⎠ = 1
⎜⎜ ⎞
⎝
⎛ p
ka
,⎟⎟ ⎠ = 1
⎜⎜ ⎞
⎝
⎛ p
kb
, assume that solutions ofx
k≡ a (mod p )
are)
1
(mod p x
x ≡
,solutions ofx
k≡ b (mod p )
arex ≡ x
2(mod p )
,then
( x
1x
2)
k= x
1kx
2k≡ ab (mod p )
, that isx
k≡ ab (mod p )
also has solution, thatis
= 1
⎟⎟ ⎠
⎜⎜ ⎞
⎝
⎛ p
kab
. Here k k
p
kab p
b p
a ⎟⎟ ⎠
⎜⎜ ⎞
⎝
= ⎛
⎟⎟ ⎠
⎜⎜ ⎞
⎝
⎟⎟ ⎛
⎠
⎜⎜ ⎞
⎝
⎛
is tenable.
If
1 ,
1 ⎟⎟ = −
⎠
⎜⎜ ⎞
⎝
= ⎛
⎟⎟ ⎠
⎜⎜ ⎞
⎝
⎛
k
k
p
b p
a
, assume the solution to
x
k≡ a (mod p )
isx ≡ x
1(mod p )
, if)
(mod p ab
x
k≡
has solution, assume the solution isx ≡ x
2(mod p )
, that is)
2
ab (mod p
x
k≡
. We saya−1as the solution to x a⋅ ≡1(mod )p . Therefore1 1 1
1 1 1 1 1
1≡(x− ⋅x)k =(x− )k⋅xk ≡ ⋅a (x− ) (mod )k p ,
1 1
(x1− )k ≡a− (mod )p . We can know
from
x
2k≡ ab (mod p )
that a−1⋅x2k ≡a ab−1 ≡b(mod )p. Because there is no solution to (mod )
xk ≡b p
, so there is also no solution to
1
2(mod )
k k
x ≡a− ⋅x p
. Neither is
1 1
(x2− ⋅x)k ≡a− (mod )p or
1 1
2 1 (mod )
x− ⋅ ≡x x− p . But we have It is a contradiction
) (mod p ab
x
k≡
, Hence⎟⎟ = − 1
⎠
⎜⎜ ⎞
⎝
⎛ p
kab
. Hence k k
p
kab p
b p
a ⎟⎟
⎠
⎜⎜ ⎞
⎝
= ⎛
⎟⎟ ⎠
⎜⎜ ⎞
⎝
⎟⎟ ⎛
⎠
⎜⎜ ⎞
⎝
⎛
.
If
1 ,
1 ⎟⎟ = −
⎠
⎜⎜ ⎞
⎝
− ⎛
⎟⎟ =
⎠
⎜⎜ ⎞
⎝
⎛
k
k
p
b p
a
, we cannot decide whether
= 1
⎟⎟ ⎠
⎜⎜ ⎞
⎝
⎛ p
kab
or
− 1
⎟⎟ =
⎠
⎜⎜ ⎞
⎝
⎛ p
kab
. For
example:if
k = 3
andp = 19
,we can know19 1 , 4 19 1
2
3 3
−
⎟ =
⎠
⎜ ⎞
⎝
− ⎛
⎟ =
⎠
⎜ ⎞
⎝
⎛
and
3 3
3
3
19
4 19
1 2 19
8 19
4
2 ⎟
⎠
⎜ ⎞
⎝
× ⎛
⎟ ⎠
⎜ ⎞
⎝
= ⎛
⎟ =
⎠
⎜ ⎞
⎝
= ⎛
⎟ ⎠
⎜ ⎞
⎝
⎛ ×
,in this condition k k
p
kab p
b p
a ⎟⎟ ⎠
⎜⎜ ⎞
⎝
= ⎛
⎟⎟ ⎠
⎜⎜ ⎞
⎝
⎟⎟ ⎛
⎠
⎜⎜ ⎞
⎝
⎛
is
tenable. But we can also know
19 1 , 5 19 1
2
3 3
−
⎟ =
⎠
⎜ ⎞
⎝
− ⎛
⎟ =
⎠
⎜ ⎞
⎝
⎛
and
3 3
3
3
19
4 19
1 2 19
10 19
5
2 ⎟
⎠
⎜ ⎞
⎝
× ⎛
⎟ ⎠
⎜ ⎞
⎝
≠ ⎛
−
⎟ =
⎠
⎜ ⎞
⎝
= ⎛
⎟ ⎠
⎜ ⎞
⎝
⎛ ×
,in this condition k k
p
kab p
b p
a ⎟⎟
⎠
⎜⎜ ⎞
⎝
= ⎛
⎟⎟ ⎠
⎜⎜ ⎞
⎝
⎟⎟ ⎛
⎠
⎜⎜ ⎞
⎝
⎛
is not
tenable. Hence, k k
p
kab p
b p
a ⎟⎟
⎠
⎜⎜ ⎞
⎝
= ⎛
⎟⎟ ⎠
⎜⎜ ⎞
⎝
⎟⎟ ⎛
⎠
⎜⎜ ⎞
⎝
⎛
is not always tenable.
2. The solvability of
x
k≡ a (mod p )
If
k
is an odd primeIf
k
is an odd prime, we have( p − x )
k≡ ( − x )
k= − x
k(mod p )
. That means the first half of complete residue system modulop
is corresponding to the second half of complete residue system modulop
. So we only need to discuss the first half part of complete residue system modulop
here.
2.1.1 If
k = 3
We have the following conclusions
Theorem 3 If
p ≡ − 1 (mod 3 )
, there is always1
3
⎟⎟ =
⎠
⎜⎜ ⎞
⎝
⎛ p
a
, ifp ≡ 1 (mod 3 )
, there areonly
3
− 1
p
a
in the complete residue system modulop
make1
3
⎟⎟ =
⎠
⎜⎜ ⎞
⎝
⎛ p
a
tenable.Proof:If
p ≡ 1 (mod 3 )
, according to Euler criterion, for everya
, There are3
solutions tox
3≡ a (mod p )
. For differenta
, the solution could not be the same. Hence there are only3
− 1
p
a
make1
3
⎟⎟ =
⎠
⎜⎜ ⎞
⎝
⎛ p
a
in a complete residue system modulop
.If
p ≡ − 1 (mod 3
,we have( 3 , p − 1 ) = 1
. Therefore there is a solution to1 ) 1 (
3 u + v p − =
. We say them asu
0,v
0.
Then
a
1= a
3u0+v0(p−1)= ( ) ( ) a
u0 3⋅ a
v0 p−1≡ ( ) a
u0 3(mod p )
. Hence, there always exists solution tox
3≡ a (mod p )
. Thus the theorem is tenable.Thus, theorem 3 is tenable.
If
p ≡ 1 (mod 3 )
,we also have following conclusions:
Corollary 1 If
x
1, x
2 satisfyx
13≡ 1 (mod p )
,x
23≡ 1 (mod p )
, and also)
\
2(mod
1
x p
x ≡
,x
1≡ \ 1 (mod p )
,x
2≡ \ 1 (mod p )
, thenx
12≡ x
2(mod p )
. Proof :It is easy to know from above that,x
1,x
2 are two different solutions to) (mod 0
2
1
p x
x + + ≡
. Fromx
13≡ 1 (mod p )
we have( x
12)
3= ( x
13)
2≡ 1 (mod p )
So
x
12also satisfyx
13≡ 1 (mod p )
. Also, becausex
2+ x + 1 ≡ 0 (mod p )
only has twosolutions,
x
12≡ 1 (mod p ), x
12≡ x
1(mod p )
are all impossible. Hence,x
12≡ x
2(mod p )
.Corollary 2 If
x
satisfyx
3≡ 1 (mod p )
, and alsox ≡ \ 1 (mod p )
, then)
(mod 1 ) 1
( x +
3≡ − p
, and also( x − 1 )
3≡ ( x + 2 )
3(mod p )
. Proof :Ifx
satisfiesx
3≡ 1 (mod p )
andx ≡ \ 1 (mod p )
,x
satisfies) (mod 0
2
1
p x
x + + ≡
, so( x + 1 )
3= x
3+ 3 x
2+ 3 x + 1 ≡ 1 + 3 ( − 1 ) + 1 = − 1 (mod p )
Also( x + 2 )
3− ( x − 1 )
3= ( x + 2 − x + 1 )( x
2+ 4 x + 4 + x
2+ x − 2 + x
2− 2 x + 1 )
) (mod 0 ) 3 3 3 (
3 x
2+ x + ≡ p
=
, that is( x − 1 )
3≡ ( x + 2 )
3(mod p )
Corollary 3 If there is
x
13≡ 1 (mod p )
,x
23≡ 1 (mod p )
, andx
1≡ \ x
2(mod p )
,)
(mod
\ 1
1
p
x ≡
,x
2≡ \ 1 (mod p )
,1 ≤ x
1< x
2≤ p − 1
, thenx
1+ x
2= p − 1
. Proof :From Corollary 1, there isx
1+ x
2≡ x
1+ x
12≡ − 1 (mod p )
.If
2
,
21
1
≥ p − x
x
, then according to Corollary 2,( x
1+ 1 )
3≡ − 1 (mod p )
,)
(mod 1 ) 1
( x
2+
3≡ − p
.Because2 , 1 1 , 1
21
> − +
+ p
x
x
, so2 , ) 1 1 ( ), 1
(
1 2−
≤ +
− +
− p
x p x
p
also( p − ( x
1+ 1 ) )
3≡ 1 (mod p ), ( p − ( x
2+ 1 ) )
3≡ 1 (mod p )
, this is contradictory to) (mod 0
2
1
p x
x + + ≡
only has solutionsx
1, x
2, and2 ,
21
1
≥ p − x
x
, thus there is2 , 1 ≤
1< p − 1
x 1 < x
2≤ p − 1
, Thus2 ) 1 ( 1 3 2
1
2 1
= −
−
− +
<
+ p
p p x
x
, thusx
1+ x
2= p − 1
.
2. 1. 2 If
k
is an odd prime andk > 3
We can reach similar conclusions:Theorem 4 If
p ≡ \ 1 (mod k )
,⎟⎟ ⎠ = 1
⎜⎜ ⎞
⎝
⎛ p
ka
is tenable; ifp ≡ 1 (mod k )
, there arek p − 1
a
in complete residue system modulop
make⎟⎟ = 1
⎠
⎜⎜ ⎞
⎝
⎛ p
ka
tenable.Proof :If
p ≡ 1 (mod k )
, according to Euler criterion, for everya
, There arek
solutions tox
k≡ a (mod p )
. For differenta
, the solution could not be the same. Hence there areonly
k p − 1
a
make= 1
⎟⎟ ⎠
⎜⎜ ⎞
⎝
⎛ p
ka
in a complete residue system modulo
p
.If
p ≡ \ 1 (mod k )
,we have( k , p − 1 ) = 1
. Therefore there is a solution to1 ) 1 ( − = + v p
uk
. We say them asu
0,v
0.
Then
a
1= a
u0k+v0(p−1)= ( ) ( ) a
u0 k⋅ a
v0 p−1≡ ( ) a
u0 k(mod p )
. Hence, there always exists solution tox
k≡ a (mod p )
. Thus the theorem is tenable.We can also popularize this Corollary:
Corollary 4 If
x
i(i=1, 2,…,k−1) meet
x
ik≡ 1 (mod p )
(i=1, 2,…,k−1) and
) (mod
\ 1 p
x
i≡
.
∀ i ≠ j
(i j, =1, 2,…,k−1),
x
i≡ \ x
j(mod p )
. Then
j
∀ i,
(i j, =1, 2,…,k−1),
∃ n
(n=1, 2,…,k−1),x x (mod p )
j n i
≡
∋
.Proof :It is easy to know from above that
x
i are thek − 1
solutions to) (mod 0
1
p x
k
i i
k
≡
∑
=− . From
x
ik≡ 1 (mod p )
there is( x
i2)
k= ( x
ik)
2≡ 1 (mod p )
,) (mod 1 ) ( )
( x
i3 k= x
ik 3≡ p
, ……,( x
ik−1)
k= ( x
ik)
k−1≡ 1 (mod p )
, so1 3
2
,
i, … … ,
ik−i
x x
x
also satisfyx
k≡ 1 (mod p )
. Also because0 (mod )
1
p x
k
i i
k
≡
∑
=− only has
− 1
k
solutions,∀ i ∈ [ 2 , k − 1 ]
andi ∈ Z
,x
1i≡ 1 (mod p )
and to∀ i , j ∈ [ 2 , k − 1 ]
,Z j
i , ∈
andi ≠ j
,x
1i≡ x
1j(mod p )
are all impossible, thusj
∀ i,
(i j, =1, 2,…,k−1),
∃ n
(n=1, 2,…,k−1),∋ x
in≡ x
j(mod p )
.
Corollary 5 If
x
i(i=1, 2,…,k−1) meet
x
ik≡ 1 (mod p )
(i=1, 2,…,k−1)and
) (mod
\ 1 p
x
i≡
.
∀ i ≠ j
(i j, =1, 2,…,k−1),
x
i≡ \ x
j(mod p )
. then we have
∑
−=
−
1
≡
1
) (mod 1
k
i
i
p
x
,
∑
−= 1
≡
1 ,
) (mod 1
k
j i
j
i
x p
x
, … ,
1 1
1 1
1(mod )
k k i
i
j j
x x p
−
− =
=
∏
≡ −∑
,
) (mod 1
1
1
p x
k
i i
≡
∏
−= .
Proof :It is easy to know from above that
x
i are thek − 1
solutions to) (mod 0
1
p x
k
i i
k
≡
∑
=−
. Therefore
1 1
1 2 3
1 1 , 1
1
( )
k k
k k k
i i i j
i i j k
i
x x x x x x x x
− −
− − −
= ≤ ≤ −
=
− = −
∑
⋅ +∑
⋅∏
-…
1 1
1 1
1
1 1 1
(mod )
k
i k
k k
i i
i
j j i i
x
x x x p
x
−
−
− −
=
= = =
+ ⋅ ≡
∑ ∏ ∏ ∑
-
. Hence
∑
−=
−
1
≡
1
) (mod 1
k
i
i
p
x
,
∑
−= 1
≡
1 ,
) (mod 1
k
j i
j
i
x p
x
, … ,
1 1
1 1
1(mod )
k k i
i
j j
x x p
−
− =
=
∏
≡ −∑
,
) (mod 1
1
1
p x
k
i i
≡
∏
−= .
2. 2 If
k = 2
nIf
k = 2
n, because( p − x )
2n≡ ( − x )
2n= x
2n, that is the first half of complete residue system of2
n degree modulop
is corresponding to the second half of complete residuesystem of
2
n degree modulop
. So we only need to discuss the first half part of complete residue system modulop
here.So we only need to discuss the first half part of complete residue system of
2
n degree modulop
here.
1. 2. 1 If
n = 1 , k = 2
This is the quadratic residue that we are familiar with. Theories of it is already quite perfect, here we will add some more.
Theorem 5 :To
∀
x
1, x
2,x
1≡ \ x
2(mod p )
, and2 1 ≤
1<
2≤ p − 1
x
x
, there is)
\
22(mod
2
1
x p
x ≡
. Ifp ≡ − 1 (mod 4 )
, to∀ x
1, x
2,x
1≡ \ x
2(mod p )
,and also, we allhave
x
12≡ \ − x
22(mod p )
. Ifp ≡ 1 (mod 4 )
, to∀ x
1,∃ x
2, and alsox
1≡ \ x
2(mod p )
, makex
12≡ − x
22(mod p )
tenable.Proof :Assume
∃ x
1, x
2, and alsox
1≡ \ x
2(mod p )
, makex
12≡ x
22(mod p )
tenable, thenp | x
22− x
12= ( x
2+ x
1)( x
2− x
1)
, take1 ≤ x
1< x
2≤ p − 1
,then
1 ≤ x
2− x
1< p − 1
,1 ≤ x
2+ x
1< 2 ( p − 1 )
, sox
2+ x
1= p
. Also because2 2
2
( )
)
( p − x = − x = x
, sop = 2 x
2= 2 x
1, this is contradictory to thatp
is an odd prime. Hence, there is nox
1, x
2 ,x
1≡ \ x
2(mod p )
, makex
12≡ x
22(mod p )
tenable.Assume
∃ x
1, x
2, andx
1≡ \ x
2(mod p )
, makex
12≡ − x
22(mod p )
tenable, thenp | x
22+ x
12, becausep
is an odd prime,p ≡ 1 (mod 4 )
; on the contrary, if) 4 (mod
≡ 1
p
,0 (mod 2 ) 2
1 ≡
−
p
. If there is⎟⎟ = 1
⎠
⎜⎜ ⎞
⎝
⎛ p
a
, then there is) (mod 1
2 1
p a
p
≡
−
, and also
( )
21 (mod )
1 2
1
p a
a
p p
≡
=
−
−
−
, that is
⎟⎟ = 1
⎠
⎜⎜ ⎞
⎝
⎛ − p
a
. Proof isfinished.
From above we can know that , if
p ≡ − 1 (mod 4 )
, absolute value of quadratic residue modulo p can range through the first part of complete residue system modulo p , but theopposite numbers are not its residue at the same time . And if
p ≡ 1 (mod 4 )
, absolute value of quadratic residue modulo p cannot range through the first part of complete residue systemmodulo p , but the opposite numbers are among residue modulop
. Because1
2
⎟⎟ =
⎠
⎜⎜ ⎞
⎝
⎛ p
n
, we can have the following corollary:
Corollary 6 If
p ≡ − 1 (mod 4 )
,1
2
⎟⎟ = −
⎠
⎜⎜ ⎞
⎝
⎛ − p
n
. Ifp ≡ 1 (mod 4 )
时,1
2
⎟⎟ =
⎠
⎜⎜ ⎞
⎝
⎛ − p
n
. thatis 2
2 1
) 1 (
−
−
⎟⎟ =
⎠
⎜⎜ ⎞
⎝
⎛ −
pp
n
. Specially, ifn = 1
, there is 21
) 1 1 ⎟⎟ = ( −
−⎠
⎜⎜ ⎞
⎝
⎛ −
pp
.
Theorem 6 If
p ≡ − 1 (mod 4 )
,(mod ) ( ) 1 4
1 ⎟⎟ ⎠ =
⎜⎜ ⎞
⎝
⇒ ⎛
⎟ ∈
⎠
⎜ ⎞
⎝
≡ ⎛ +
p N a
n p p
a
n
;if
) 4 (mod
≡ 1
p
,p
na ⎟
⎠
⎜ ⎞
⎝
≡ ⎛ − 4
1
or(mod )
4
1 p
p
n⎟ ⎠
⎜ ⎞
⎝
− ⎛ − ( n ∈ N ) ⎟⎟ ⎠ = 1
⎜⎜ ⎞
⎝
⇒ ⎛ p
a
.Proof :If
p ≡ − 1 (mod 4 )
,(mod )
4 ) 1 1 4 (
1 2
1
2p p p p
p ⎟ = + + ≡ +
⎠
⎜ ⎞
⎝
⎛ +
, thatis
(mod )
4
2
1 p p
x +
≡
must have solution. If⎟⎟ = 1
⎠
⎜⎜ ⎞
⎝
⎛ p
a
,⎟⎟ = 1
⎠
⎜⎜ ⎞
⎝
= ⎛
⎟⎟ ⎠
⎜⎜ ⎞
⎝
⎛
n np a p
a
, thatis
p
n⎟ ⎠
⎜ ⎞
⎝
⎛ + 4
1 ( n ∈ N
∗)
is also quadratic residue modulop
, also 1 is quadratic residue modulop
, thusp
n⎟ ⎠
⎜ ⎞
⎝
⎛ + 4
1 ( n ∈ N )
are all the quadratic residue modulop
.If
p ≡ 1 (mod 4 )
,(mod )
4 ) 1 1 4 (
1 2
1
2p p p p
p ⎟ = − − ≡ − −
⎠
⎜ ⎞
⎝
⎛ −
, thatis
(mod )
4
2
1 p p
x ≡ − −
must has solution. According to Theorem1. 1. 1(2),) 4 (mod
2
1 p p
x ≡ −
also has solution. If⎟⎟ ⎠ = 1
⎜⎜ ⎞
⎝
⎛ p
a
,⎟⎟ ⎠ = 1
⎜⎜ ⎞
⎝
= ⎛
⎟⎟ ⎠
⎜⎜ ⎞
⎝
⎛
n np a p
a
, thatis
n
n
p
p ⎟
⎠
⎜ ⎞
⎝
− ⎛ −
⎟ ⎠
⎜ ⎞
⎝
⎛ −
4 , 1 4
1 ( n ∈ N
∗)
is also quadratic residue modulop
, also ,1, ‐1 is quadratic residue modulop
. Thus,n
n
p
p ⎟
⎠
⎜ ⎞
⎝
− ⎛ −
⎟ ⎠
⎜ ⎞
⎝
⎛ −
4 , 1 4
1 ( n ∈ N )
are all the quadratic residue ofp
.But whether
p p ( n N )
a
n
⎟ ∈
⎠
⎜ ⎞
⎝
≡ ⎛ + (mod ) 4
1
、p
na ⎟
⎠
⎜ ⎞
⎝
≡ ⎛ − 4
1
or) 4 (mod
1 p
p
n⎟ ⎠
⎜ ⎞
⎝
− ⎛ − ( n ∈ N )
is also the necessary condition of⎟⎟ ⎠ = 1
⎜⎜ ⎞
⎝
⎛ p
a
when) 4 (mod
− 1
≡
p
、p ≡ 1 (mod 4 )
时⎟⎟ ⎠ = 1
⎜⎜ ⎞
⎝
⎛ p
a
, this is still a conclusion that we have notproved in this paper.
Theorem 7
3 ⎟⎟ = 1 ⇔ ≡ 1
⎠
⎜⎜ ⎞
⎝
⎛ p
p
or− 1 (mod 12 )
,5 ⎟⎟ = 1 ⇔ ≡ 1
⎠
⎜⎜ ⎞
⎝
⎛ p
p
or− 1 (mod 10 )
,1 7 1
≡
⇔
⎟⎟ =
⎠
⎜⎜ ⎞
⎝
⎛ p
p
or− 1
,3
or− 3
,9
or− 9 (mod 28 )
.Proof :If
a = 3
,we can know) 3 (mod 1 3 ⎟ = 1 ⇔ ≡
⎠
⎜ ⎞
⎝
⎛ p p
. According to law of quadratic
reciprocity,
2 1 2
1 2
1 3
) 1 ( )
1 3 (
3
−
⋅ −
−
−
=
−
⎟⎟ =
⎠
⎜⎜ ⎞
⎝
⎟ ⎛
⎠
⎜ ⎞
⎝
⎛
p pp p
. Therefore
⎪ ⎪
⎩
⎪⎪ ⎨
⎧
− ≡
⎟ =
⎠
⎜ ⎞
⎝
⎛
⇔
⎟⎟ =
⎠
⎜⎜ ⎞
⎝
⎛
) 2 (mod 2 0
1 3 1 3 1
p p p
or
⎪ ⎪
⎩
⎪⎪ ⎨
⎧
− ≡
−
⎟ =
⎠
⎜ ⎞
⎝
⎛
) 2 (mod 2 1
1 3 1 p
p
.
⎩ ⎨ ⎧
≡
≡
) 4 (mod 1
) 3 (mod 1 p p
or
⎩ ⎨ ⎧
−
≡
−
≡
) 4 (mod 1
) 3 (mod 1 p p
. Hence
p ≡ 1 (mod 12 )
or)
12 (mod
− 1
≡
p
.If
a = 5
,we can know) 5 (mod 1 5 ⎟ = 1 ⇔ ≡
⎠
⎜ ⎞
⎝
⎛ p p
or
p ≡ − 1 (mod 5 )
. According to law ofquadratic reciprocity,
1 )
1 5 (
5
2 1 2
1 5
=
−
⎟⎟ =
⎠
⎜⎜ ⎞
⎝
⎟ ⎛
⎠
⎜ ⎞
⎝
⎛
−⋅p−p p
. Therefore
5 1 5 1
⎟ =
⎠
⎜ ⎞
⎝
⇔ ⎛
⎟⎟ =
⎠
⎜⎜ ⎞
⎝
⎛ p
p
.) 5 (mod
≡ 1
p
orp ≡ − 1 (mod 5 )
. In addition,p ≡ 1 (mod 2 )
. Hence,p ≡ 1 (mod 10 )
or)
10 (mod
− 1
≡
p
.If
a = 7
, we can know) 7 (mod 1 7 ⎟ = 1 ⇔ ≡
⎠
⎜ ⎞
⎝
⎛ p p
or
p ≡ − 3 (mod 7 )
orp ≡ 2 (mod 7 )
.According to law of quadratic reciprocity,
2 1 2
1 2
1 7
) 1 ( )
1 7 (
7
−
⋅ −
−
−
=
−
⎟⎟ =
⎠
⎜⎜ ⎞
⎝
⎟ ⎛
⎠
⎜ ⎞
⎝
⎛
p pp p
. Therefore
⎪ ⎪
⎩
⎪⎪ ⎨
⎧
− ≡
⎟ =
⎠
⎜ ⎞
⎝
⎛
⇔
⎟⎟ =
⎠
⎜⎜ ⎞
⎝
⎛
) 2 (mod 2 0
1 7 1 7 1
p p p
or
⎪ ⎪
⎩
⎪⎪ ⎨
⎧
− ≡
−
⎟ =
⎠
⎜ ⎞
⎝
⎛
) 2 (mod 2 1
1 7 1 p
p
.