Complete Solutions 7.

4 Page 1 of 25

Complete Solutions to Exercise 7.4

1. In each case we use the corollary:

  q 
  
 p    p 
if p  1 mod 4   
or q  1 mod 4 
(7.17)    
 q    q 
 
  p 
if p  3 mod 4   
and q  3 mod 4 
 
(a) We are asked to see if x 2  12 mod 89 is solvable. This means we need to
 12 
determine the Legendre symbol   . We know that 12  22  3 so
 89 
 12   22   3     
         1   3    3  (†)
 89   89   89   89   89 

1 because 22 is a quadratic residue

 
Since 89  1 mod 4 so by using (7.17) on the right-hand side of (†) we have
 3   89   2 
       
 89   3   3 
 Because 89  2 mod 3
   .
Applying the test for residue 2, Proposition (7.15):

 2  
   
 1



if p  1 mod 8 
 p  1 if p  3 mod 8


2 2
 
To   with 3  3 mod 8 gives    1 . By (†) and this result we have
 3   3 
 12   3   2 
         1 .
 89   89   3 
 12 
Since    1 so 12 is a quadratic non-residue of 89 which implies that
 89 

 
x 2  12 mod 89 is unsolvable.

 
(b) We need to test whether x 2  40 mod 101 is solvable. This means we must
 40 
find the Legendre symbol   . The prime decomposition of 40 is
101
40  8  5  23  5 :
 Complete Solutions 7.4 Page 2 of 25

 40   23  5   23   5 
      
101  101  101 101
 22   2   5 
        
 101  101 101

1
 2   5 
      †
101 101
We use our normal test for residue 2, Proposition (7.15):

 2  
   
 1
 
if p  1 mod 8 
 p  1 if p  3 mod 8



 2 

Since p  101  5  3 mod 8 so  101
  1 . Evaluating the other term in (†):

 5  101  1 
         1 .
 
101 Because 51 mod 4  5  Because 1011 mod 5  5 
   
[Remember 1 is a quadratic residue of any odd prime p.]
 2  1
Substituting    1 and    1 into (†) gives
101  5 
 40   2   5 
101  101  101  1  1  1 .
     


Hence 40 is a quadratic non – residue of 101 so x 2  40 mod 101 cannot be solved. 

(c) We are given the quadratic congruence x 2  36 mod 1223 and since 36  62 so 
  
x  6 mod 1223 . Therefore x 2  36 mod 1223 is solvable. 
(d) We have to find whether x 2  89 mod 197 is solvable. This means we need to
 89 
calculate the Legendre symbol  
 . Since 89  1 mod 4 so by (7.17) we have
197  
 Complete Solutions 7.4 Page 3 of 25

 89  197   19 
     
197    89    89 
 Because 197  19 mod 89 
   
 89 
  
 19    
 Because 89  1 mod 4 so applying (7.17)

13 
  
 19 
Because 89  13 mod 19 


 

19 
  
13    
Because 13  1 mod 4 so applying (7.17)

6
  
13
   
Because 19  6 mod 13 
   
2 3
       
Because 6  2  3
13  13 
So far we have
 89   2   3 
     
197   13   13  (‡)

The residue 2 is tested by using Proposition (7.15):

 2  
   
 1



if p  1 mod 8 
 p  1 if p  3 mod 8


2
 
Since 13  5  3 mod 8 so using this proposition we have    1 .
13 

Evaluating the other Legendre symbol on the right - hand side of (‡):
 3  13 
    
13   3 
 Because 13  1 mod 4
   so applying (7.17)

1
    1
 3 
 Because 13  1 mod 3 
   
2 3
Putting these    1 and    1 into (‡) gives
13  13 
 89   2   3 
     
197   13   13   1  1  1 .

 89 
Since  
  1 so x 2  89 mod 197 is not solvable.
197 

197 
 
(e) We need to test whether x 2  197 mod 89 is solvable. We have to find   .
 89 
197   89 
 
Since 89  1 mod 4 so by (7.17) we have     .
 89  197 
 Complete Solutions 7.4 Page 4 of 25

 89  197 
This was evaluated in part (d) and we had    1    . Hence
197   89 


x 2  197 mod 89 is unsolvable. 

 p 1 q 1

2. We need to find 1    

  
 2   2 
for (i) p  1 mod 4 , q  3 mod 4 and


(ii) p  3 mod 4 , q  1 mod 4 .   

(i) We are given that p  1 mod 4 , q  3 mod 4 so there are integers k and m  
such that
p  4k  1 and q  4m  3
 p  1 q  1
Substituting these into the index      gives
 2   2 
 p  1 q  1  4k  1  1  4m  3  1
          
 2   2   2   2 
 2k  2m  1 Even Number
 
 p 1 q 1
2k2m 1
So 1  1
  
 2   2 
1 .

 
(ii) Similarly, for p  3 mod 4 , q  1 mod 4 by interchanging p and q we have  
 p 1 q 1

1
 
 2  2 
 1.

3. We need to show that:

 p   q  
       
 1 if p  1 mod 4 or q  1 mod 4
   
 q   p  1 if p  q  3 mod 4


  
Proof.
We have by the Law of Quadratic Reciprocity:
 p   q   p 1 q 1
      1 2  2 
 q   p 
 p   q   p 1 q 1

So       1 2   2  .
    
  

 q   p 
We found in Example 17 and question 2 above that if p  1 mod 4 or  
 p   q   p1 q 1

 
q  1 mod 4 then       1 2   2   1 and if p  q  3 mod 4 then
   
 
 

 q   p 
 Complete Solutions 7.4 Page 5 of 25

 p   q   p1 q 1

      1 2  2   1 .
 q   p 
This is our required result.
■

 p

   if p  1 mod 4
 3  

  3   

4. We need to prove for p  3 that     .
 p    p 


   if p  3 mod 4
  
 3


Proof.
To prove this we use the following corollary:
  q 
 p    p 
 if p  1 mod 4  
or q  1 mod 4 
(7.17)    
 q    q 
 
  p 
if p  3 mod 4  and q  3 mod 4  

 
If our given p  1 mod 4 then by this corollary we have
 3   p 
     .
 p   3 

 
If p  3 mod 4 then by this corollary, we have
 3   
    p  .

 p   3 
This proves our required result.
■

5. (i) We are required to prove that for prime p  3 we have

 3  
   
 1 if p  1 mod 6
  
 p  1 if p  5 mod 6



 
Proof.
Since 3  1  3 so
 3  1  3 
        (†)
 p   p   p 
By (7.11) we have

 1 
   

 1 if p  1 mod 4  
 p  1 if p  3 mod 4


  
 Complete Solutions 7.4 Page 6 of 25

By result of question 11(i) of Exercise 7.3 we have

 3  
    
 1 if p  1 or 11 mod 12
 
 p  1 if p  5 or 7 mod 12



 
We use these two results and (†).

 
Consider the two cases (i) p  1 mod 6 and (ii) p  5 mod 6 .  
Case (i)

 
Let p  1 mod 6 then there is a positive integer such that p  6k  1 . Now either
k is even or odd. Let us first take k to be even then k  2m where m is a positive
integer. Substituting this into p  6k  1 gives
p  6 2m   1  12m  1  4 3m   1 .

   
Hence p  1 mod 12 and p  1 mod 4 so using the above results and (†):
 3   1  3 
         1  1  1 .
 p   p   p 
Now let us take k to be odd so k  2n  1 where n is a positive integer. Putting
this into p  6k  1 gives
p  6 2n  1  1  12n  7  4 3n  1  3 .

   
Therefore p  7 mod 12 and p  3 mod 4 so again using the above results with
(†):
 3   1  3 
         1  1  1 .
 p   p   p 
 3 
 
In both cases (k is odd and even) we have    1 if p  1 mod 6 . This proves
 p 
the first part of the result.
Case (ii)

 
Let p  5 mod 6 then there is a positive integer k such that p  6k  5 . Now
either k is even or odd. Let us first take k to be even then k  2m where m is a
positive integer. Substituting this into p  6k  5 gives
p  6 2m   5  12m  5  4 3m  1  1 .

   
Hence p  5 mod 12 and p  1 mod 4 so using the above results and (†):
 3   1  3 
         1  1  1 .
 p   p   p 
 Complete Solutions 7.4 Page 7 of 25

Now let us take k to be odd so k  2n  1 where n is a positive integer. Putting

this into p  6k  5 gives
p  6 2n  1  5  12n  11  4 3n  2  3 .

   
Therefore p  11 mod 12 and p  3 mod 4 so again using the above results with
(†):
 3   1  3 
         1  1  1 .
 p   p   p 
 3 
 
In both cases (k is odd and even) we have    1 if p  5 mod 6 . This proves
 p 
the second part of the result.
 3  
Therefore, we have    


 1 if p  1 mod 6

 
. This is our required result.
 p  1 if p  5 mod 6


 
■
(ii) We use the result of part (i) to factorize each of the integers in this part.
(a) We are asked to find the prime factorization of 1042  3  10 819 . Since our
integer is of the form n 2  3 so the odd prime factors p of this 1042  3 must

 
satisfy p  1 mod 6 . The first few primes are 7, 13, 19, 31, 37, 43, ---. Dividing
10 819 by each of these we find that
10 819  31  349 .

We have  349   18 and none of the primes in the above list below 18 go into 349,
 
so 349 is prime. Hence 10 819  31  349 .
(b) We need to find the prime factorization of 2362  3  55 699 . Let p be a prime

 
factor of 2362  3 then p  1 mod 6 and the primes of this form are 7, 13, 19, 31,
37, 43, 61, 67, 73 --- and we find that
55 699
 763  55 699  73  763 .
73
We just need to factorize 763 but let us first see which primes we need to test. We
have  763   27 and there are no primes in the above list which are below 27 and
 
go into 763. How do we know this?
Because if there was a smaller prime then it would also be a factor of 55 699 and
the first prime to be a factor of 55 699 is 73. Therefore 763 is prime and
55 699  73  763 .
 Complete Solutions 7.4 Page 8 of 25

(c) We are asked to factorize 3622  3  131 047 . Let p be a prime factor of this

 
number then p  1 mod 6 . The first of these is 7 and we find that
131 047
 18 721 .
7
From this we have  18 721   136 . We need to now try prime factors which satisfy
 
 
p  1 mod 6 and first few are 7, 13, 19, 31, 37, 43, 61, 67, 73, 79 and 97 which is a
factor of 18 721 because 18 721  97  193 . Now 193 is prime so the prime
factorization of 131 047  7  97  193 .

6. We are required to prove that prime factors of the integer n 2  n  1 are of the
form 6k  1 .
Proof.
Let p be an arbitrary prime factor of the given integer n 2  n  1 . We have


n 2  n  1  0 mod p . 
By using the given hint in the question consider the integer

2n  1  
2
 4n 2  4n  1  4 n 2  n  1  3 .
Using the first result we have

2n  1    
2
 4 n 2  n  1  3  0  3  3 mod p (*)

0 mod p 

Let x  2n  1 and substituting this into (*) yields


x 2  3 mod p 
This is a quadratic congruence. Hence we need to find for which primes p is the
 3 
Legendre symbol    1 because when the Legendre symbol is equal to 1 we have
 p 
a quadratic residue.
By the result of the previous question:

 3  
   
 1 if p  1 mod 6
  
 p  1 if p  5 mod 6


  

Hence the prime p must be of the form p  1 mod 6 so p  6k  1 for some 
positive integer k.
■
 Complete Solutions 7.4 Page 9 of 25

7. We are asked to prove that there are infinitely many primes of the form 8k  1 .
Proof.
We use the given hint and suppose there are finitely many primes of the form
8k  1 which we can denote as p1, p2 , , pk . Let n  p1  p2    pk and

consider the integer N  4n   2 . Clearly N is composite because 2 is factor of N.

2

Let p be a prime factor greater than 2 (or an odd prime) of N. Then

N  4n   2  0 mod p .  
2

Let x  4n then we have

N  x 2  2  0 mod p   x 2  2 mod p  
2
 
This x 2  2 mod p and is solvable because x  4n so    1 . By question 3(i) of
 p 
Exercises 7.3:

2 is a quadratic residue of prime p  p  1 mod 8 .  

 
we have p  1 mod 8 , which implies that
p  8k  1 or p  8k  1 .
If all the prime factors greater than 2 of N are of the form 8k  1 then the product
of this is also of the form 8m  1 (this can be shown by induction) but this is

impossible because N  4n   2 . Hence N must have a prime factor of the form
2

p  8k  1 . Clearly this p  8k  1 is not one in the above list p1, p2 , , pk . Why

not?

4n  and p N and since N  4n   2 so

2 2
If it is then p n so p

p 2.
This is impossible because p is an odd prime.
Hence there are infinitely many primes of the form 8k  1 .
■

q 1 2  p 1 2 
k p k q 
8. We need to find the given sum  
 q

 p

 
 for the primes p  17
k 1 
k 1  
p 1 q 1
and q  13 . Substituting these p  17 and q  13 into and
2 2
respectively gives
 Complete Solutions 7.4 Page 10 of 25

p  1 17  1 q  1 13  1
  8 and   6.
2 2 2 2
Evaluating the sum separately:
6 
k  17   1  17   2  17   3  17   4  17   5  17   6  17 
  13    13    13    13    13    13    13 
k 1              
 17   34   51   68   85   102 
      
 13   13   13   13   13   13 
           
 1  2  3  5  6  7  24
8                  
  k  13    1  13    2  13    3  13    4  13    5  13    6  13    7  13    8  13 

k 1  17
  17   17   17   17   17   17   17   17 
                
 13   26   39   52   65   78   91   104 
        
 17   17   17   17   17   17   17   17 
               
 0  1  2  3  3  4  5  6  24

Adding both these summations gives

6 
k  17  8 
k  13 
  13    17   24  24  48
 
k 1   k 1  
Drawing the graph gives:

The number of lattice points shown in this graph is 6  8  48 which is given by

the above sum.

9. (See Example 17.) Consider the two different cases:


(a) p  1 mod 4  
or q  1 mod 4 . 
 Complete Solutions 7.4 Page 11 of 25

(b) Both p  3 mod 4   and q  3 mod 4 .  

Case (a)

 
Let p  1 mod 4 then there exists a positive integer k such that p  4k  1 .

 p  1 q  1
Putting this into the index of Law of Quadratic Reciprocity      we
 2   2 

have
 p  1 q  1  4k  1  1 q  1 q  1
            2k    Even because q is odd
 2   2   2   2   2   
 p 1 q 1 q 1
2 k
 p   q     
2   2   2 
      1  1


1  Because index is even

 q   p   

 p  q   p  q   p  q   p  q 
Since       1 so       1 or       1 . Either way      .
 q   p   q   p   q   p   q   p 

We can present the same argument with q  1 mod 4 .  

Case (b)


Let p  3 mod 4   
and q  3 mod 4 then there are positive integers k and m such

that
p  4k  3 and q  4m  3 .
Substituting this into the index of the Law of Quadratic Reciprocity yields
 p  1 q  1  4k  3  1  4m  3  1
          
 2   2   2   2 
 4k  2   4m  2 
       2k  1  2m  1 Odd number
 2   2   

 p  q  2k 12m 1

Therefore       1  1 because 2k  12m  1 is odd.
 q   p 
p q  p q  p q 
We have    1 and    1 or    1 and    1 . Hence      .
 q   p   q   p   q   p 
This completes our proof.
■
 Complete Solutions 7.4 Page 12 of 25

p 1 2  
10. We need to prove that 
ka

   g mod 2 where p
p  a and g be the number
 
k 1

of negative residues defined in Gauss’s Lemma.

Proof.
p 1
Let S be the set of the product of k and a where k  1, 2, 3, , :
2
  p  1 
S  a, 2a, 3a, 4a, ,  a 
.

  2   

The integer g is defined as the number of negative residues in this list, these are the
p 1
ones which are greater than .
2
 p  1
We can write each of these ka as a residue of modulo p which lies between  
 2 
 p  1
and   . We can illustrate this on a modulo p clock:
 2 

Negative residue
Positive residue

Denoting each of these residues by rk, that is


ka  rk mod p . 
If rk is positive then ka is one of the least positive residues in the set

  p  1

T  1, 2, 3, 4,  ,  
 2 

 

 
 ka 
We need to consider the floor function   . So dividing these integers in the set T
p
 
by p and then evaluating the floor function gives
 1   2   3   4  
 p  1 
T '    ,   ,   ,   , , 
 p   p   p   p 
 2p  
 
 0, 0, 0, 0, , 0
  


If rk is negative, then ka is one of the negative residues in the set
 Complete Solutions 7.4 Page 13 of 25

  p  1

U  1,  2,  3,  4, ,   


  2  

 ka 
Similarly finding the floor function   gives
p
 
 1   2   3   4  
 p  1
U '    ,   ,   ,   , , 
 p   p   p   p 

 2p   
 1,  1,  1,  1, ,  1 
  


p1 2  
ka
Evaluating the sum    gives
 
k 1  p 

p1 2           
     a    2a    3a      p  1 a 
ka
 p  p  p   p   2  p 
k 1          
 0  0  1  1    1  0   0

g Number of negative residues

 g mod p 
Taking modulo 2 we have
p 1 2  

ka
   g  g
p mod 2
k 1  
This is our required result.
■

p1 2  
ka
11. Substituting p  13 and a  16 into   gives
p
 
k 1

6  16k   16   32   48   64   80   96 
  p


      
 13   13   13   13   13   13 
k 1              
 1  2  3  4  6  7  23
We have ka  16, 32, 48, 64, 80, 96 , Writing these integers as residues between
6 and 6 of modulo 13 gives
ka  16  3, 32  6, 48  4, 64  1, 80  2, 96  5 mod 13
Hence g  2 . We have 23   
 2 mod 2 . Why doesn’t Lemma (7.20) work in this
case?
Because a  16 but in the Lemma it states that let ‘a also be odd’.

12. We need to prove that odd prime divisors of the integer n 2  1 are of the form
4k  1 .
Proof.
 Complete Solutions 7.4 Page 14 of 25

Let p be an odd prime divisor of n 2  1 , that is

 
n 2  1  0 mod p implies that n 2  1 mod p .  
By question 6 of Exercises 7.1:

1 is a QR of an odd prime p  p  1 mod 4 .  

 
The quadratic congruence n 2  1 mod p has solutions so p  1 mod 4   which
implies that p  4k  1 .
■

13. We need to prove that there are an infinite number of primes of the form
3m  1 .
Proof.
Suppose there are a finite number of primes of the form 3m  1 which we can write
in a list as
p1, p2, p3 , , pn
Consider the integer

 
2
N  3  p1  p2  p3   pn  3.
N is composite because 3 is a factor of N. This implies that we must have a prime
factor, p  3 say of N. By using modular arithmetic we have

3  p  p  p    3  p  p   
2 2

1 2 3
   pn  3  0 mod p  1 2
 p3    pn  3 mod p

Let 3  p1  p2  p3    pn  x then the above can be written as

x 2  3 mod p . 
This is a quadratic congruence. We know it has solutions because
x  3  p1  p2  pn so it is solvable. By the result of question 5:

 3  
   
 1 if p  1 mod 6
  
 p  1 if p  5 mod 6



 
 
We have p  1 mod 6 so p  6k  1 . Writing this as a factor of 3 gives
p  3 2k   1
We have p  3 2k   1 which is of the form 3m  1 and is a prime factor of N, that

is p N . Since p is of the form 3m  1 so it must be one in the above finite list

p1, p2, p3 , , pn .
 Complete Solutions 7.4 Page 15 of 25

3  p  p  p 
2
From this we have p 1 2 3
   pn .

Since

 
2
N  3  p1  p2  p3   pn 3

3  p  p  p 
2
And p N and p 1 2 3
   pn so p 3 this is impossible because

p  3 . We have a contradiction so there are an infinite number of primes of the

form 3m  1 .


14. (a) We are required to find x in 25997  x mod 1993 given that 1993 is prime. 
Using Euler’s Criterion (7.5):

 
p1
a is a quadratic residue of p  a 2
 1 mod p

We have a  25  52 so clearly 5 is a quadratic residue of 1993 so by (7.5) we have

 
19931
25 2
 25996  1 mod 1993

Multiplying both sides of this result 25996  1 mod 1993 by 25 yields

25  25996  25997  25 mod 1993


Hence x  25 mod 1993 . 
(b) We are asked to find the least positive residue x in 26997  x mod 1993 . The  
prime decomposition of 26 is 2  13 . We need to check that if 2 and 13 are
quadratic residues of 1993 because
 26   2   13 
        (*)
1993  1993  1993 
For 2 we use (7.15)

 2  
   
 1
 
if p  1 mod 8 
 p  1 if p  3 mod 8



 2 

Since 1993  1 mod 8 so    1 which implies that 2 is a quadratic residue of
1993 
1993.
 13 
We need to find the other Legendre symbol   . Since 13  1 mod 4 so by
1993   
 Complete Solutions 7.4 Page 16 of 25

  q 
  
 p    p 
if p  1 mod 4  
or q  1 mod 4 
(7.17)    
 q    q 
 
  p 
if p  3 mod 4   and q  3 mod 4  
We have
 13  1993   4 
     
1993    13   13 
 Because 1993  4 mod 13 
   
 22 
    1  Because 22 is a quadratic residue
13   

Substituting these into (*) yields

 26   2   13 
         11  1 .
1993  1993  1993 
Hence 26 is a quadratic residue of 1993 so by Euler’s Criterion

 
19931
26 2
 26996  1 mod 1993 .


Therefore 26997  26 mod 1993 or x  26 mod 1993 .   
 p1 
15. We are required to prove that if p  8k  1 then p 2 2  1 .
 
 
Proof.
We are given p  8k  1 therefore p  1 mod 8 . By  
 2  
   
 1



if p  1 mod 8 
 p  1 if p  3 mod 8


2
We have    1 which implies that 2 is a quadratic residue of p. By Euler’s
 p 
Criterion (7.5):

 
p1
a is a quadratic residue of p  a 2
 1 mod p

Using this criterion with a  2 we have

   
p1 p1
2 2
 1 mod p  2 2
 1  0 mod p .
 p1 
 
p1
Since 2  1  0 mod p so we conclude that p 2 2  1 . This completes our

2

 
proof.
■
 Complete Solutions 7.4 Page 17 of 25

 2a   a 
16. We are asked to show that      provided p  1 mod 4 .
 p   p   
Proof.
Factorizing 2a  2  a so
 2a   2  a 
        (‡)
 p   p   p 


We are given that p  1 mod 4 therefore by (7.15): 
 2  
   
 1

if p  1 mod 8 
 p  1 if p  3 mod 8




 
This implies that if p  1 mod 8 which in turn implies that
p  8k  1  4 2k   1 then
 2 
   1 .
 p 
 2a   2   a  a  a 
Substituting this into (‡) gives          1       .
 p   p   p   p   p 
This completes our proof.
■
p 1
a 
 
2
17. We need to show that if p  1 mod 4 then   p   0 .
a 1

Proof.
p1
2 a  1 2 3  p  1
We need to show that   p    p    p    p        0.
2p 
a 1

From question 7(a) of the Exercise 7.1 we have the following result:

If a is a quadratic residue then p  a is a quadratic residue  p  1 mod 4 .  

Consider the set
 
 
p  1 p  1
S   1, 2, 3, , , , , p  3, p  2, p  1
 2  2 
First half
of the least positive residues modulo p

Last half of the least positive residues modulo p  
The above result claims that if 1 is a quadratic residue then so is p  1 and if 2 is a
quadratic residue so is p  2 and so on.
This implies that the quadratic residues in the above list are symmetrical. By
Proposition (7.4):
 Complete Solutions 7.4 Page 18 of 25

p 1 p 1
There are exactly quadratic residues and quadratic non-residues of p.
2 2
Since the list in S is symmetrical so half the residues in the first half, that is
p 1
1, 2, 3, , ,
2 

First half of the least positive residues modulo p

must be quadratic residues and half of these must be quadratic non-residues. By the
definition of the Legendre symbol (7.7):
 a    1 if a is a quadratic residue of p
    
 p  1 if a is a quadratic non-residue of p

Therefore
p1
2 a  1 2 3  p  1
  p    p    p    p      0
2p 
a 1

This is our required result.

■

18. We are given the following table:

Prime p 3 7 11 13 17 19 23 29 31

5 p  1 1 1 1 1 1 1 1 1

Prediction is
 5  
   
 1 if p  1 mod 5
  
 p  1 if p  2 mod 5


  
We need to prove this.
Proof.
We need to consider the four different cases:


(i) p  1 mod 5  
(ii) p  1 mod 5
(iii) p  2 mod 5 (iv) p  2 mod 5
In each case we use the popular corollary:
  q 
 p    p 
 if p  1 mod 4   or q  1 mod 4  
(7.17)    
 q    q 
 
  p 
if p  3 mod 4   and q  3 mod 4  

5 p
 
Since 5  1 mod 4 we have      .
 p   5 
 Complete Solutions 7.4 Page 19 of 25

Case (i):


Applying this to p  1 mod 5 : 
 5   p   1 
         1
 p   5   5 
Because p  1 mod 5 
   
Case (ii):


This time p  1 mod 5 so 
 5   p   1
       
 p   5   5 
 Because p  1 mod 5 
   
Using

(7.11)
 1 
   
 1



if p  1 mod 4 
 p  1 if p  3 mod 4


 5   1
We have       1 because 5  1 mod 4 .
 p   5 
 
Case (iii):
We consider the case p  2 mod 5 :  
 5   p   2 
       
 p   5   5 
Because p  2 mod 5 
 
 
Applying the following to 2 5 :  
 2  
   

 1 
if p  1 mod 8 
 p  1 if p  3 mod 8





In view of 5  3 mod 8 we have 
 5   2 
      1 .
 p   5 
Case (iv):
This time we have p  2 mod 5 so  
 5   p   2 
       
 p   5   5 
Because p  2 mod 5 
   
Therefore
 5   2   1  2 
     
 p   5 
  
 5 
 
 5   1  1  1
 
1 by (7.11) 1 by (7.15)

We have considered all four cases and shown our predicted formula.
 Complete Solutions 7.4 Page 20 of 25

■
To factorize each of the given integers we need to use our predicted formula:
 5  
    
 1 if p  1 mod 5
 
 p  1 if p  2 mod 5



 
(a) We are asked to factorize 1042  5  10 811 . Let p be a prime factor of

 
10 811 therefore p satisfies p  1 mod 5 . The first couple of primes of this

format are 11, 19 and we find that 10 811  19  569 . Also  569   23 so we only
 
need to test if 19 goes into 569 but it doesn’t so 569 is prime. Hence
10 811  19  569 .
(b) Like part (a) we must find the prime factorization of
5042  5  254 011

 
Let p be a prime factor of this number then p  1 mod 5 and testing primes of
this format 11, 19, 29, 31, 41, ---. Clearly 11 is not a factor because adding the
digits of 254 011 gives 1  1  0  4  5  2  1 and 11 1 . Trying 19 we have
254011  19  13 369

Also  13 369   115 so we need to test primes up to 115. Again 11 cannot be a

 
factor as it is not a factor of the original number. We find that 19 is also not a


factor. The next prime after 19 of the format p  1 mod 5 is 29 and 
13 369  29  461

Also 461 is prime because we have tested primes up to 29 and  461   21 . Hence
 
254011  19  29  461 .

19. We need to prove:

 7 

   1 if p  1,  3,  9 mod 28
 p  
Proof.
Arguing along similar lines to solution of previous question we have the following
cases:


p  1 mod 28 :
 Complete Solutions 7.4 Page 21 of 25

 
Since p  1 mod 28 so p  28k  1  7 4k   1  4 7k   1 which implies that

 
p  1 mod 7 and p  1 mod 4 . Applying (7.17) we have  
 7     1 
    p      1 .
 
 p  By (7.17)  7  Because p 1 mod 7  7 
 
 


p  1 mod 28 : 
 
Since p  1 mod 28 so p  28k  1  7 4k   1 which implies that

 
p  1 mod 7 and p  1  3 mod 4 :  
 7  p  
         1 †
 p   
 7  Because p1 mod 7  7 

By (7.17)

Now we use the test for residue 1 which is

(7.11)

 1 
    


 1 if p  1 mod 4
if p  3 mod 4
 p  1


As 7  3 mod 4 so applying (7.11) gives 1 7  1 . Substituting this into (†)
yields
 7   
    1  1  1 .

 p   7 


p  3 mod 28 : 
 
In view of p  3 mod 28 so p  28k  3  7 4k   3  4 7k   3 which implies

 
that p  3 mod 7 and p  3 mod 4 . Applying (7.17) we have  
 7  p  3 
         
 p  
By (7.17)
 7   
  Because p3 mod 7  7 
 7     

by (7.17)

3
7  1
             1
3 3
 
 Because 7  1 mod 3 
 
p  3 mod 28 :  
 
In view of p  3 mod 28 so p  28k  3  7 4k   3  4 7k   3 which implies

 
that p  3 mod 7 and p  3  1 mod 4 . Applying (7.17) we have  
 7     3   1  3 
    p           (*)
 
 p  By (7.17)  7  Because p 3 mod 7  7   7   7 
 
 
Evaluating each of the Legendre symbols on the right - hand side of (*).
 Complete Solutions 7.4 Page 22 of 25


As 7  3 mod 4 so by applying 
(7.11)
 1 
   
 1



if p  1 mod 4 
 p  1 if p  3 mod 4


we have 1 / 7   1 . Evaluating the second Legendre symbol in (*):
 3     
    7    1   1

 7  
 3   7 
 Because 7  1 mod 3 
   
 1  3
Substituting    1 and    1 into (*) gives
 7   7 
 7  1   3 
         1  1  1 .
 p   7   7 


p  9 mod 28 : 
 
In view of p  9 mod 28 so p  28k  9  7 4k  1  2  4 7k  2  1 which

implies that p  2 mod 7 and p  1 mod 4 . Applying (7.17) we have

 7     2 
    p    
 
 p  By (7.17)  7  Because p2 mod 7  7 
 
 
Using the test for residue 2 which is (7.15):

 2  
   
 1
 
if p  1 mod 8 
 p  1 if p  3 mod 8




 
As 7  1 mod 8 so by applying this (7.15) we have
 7   2 
      1 .
 p   7 


p  9 mod 28 : 
 
Since p  9 mod 28 so p  28k  9  7 4k  1  2  4 7k  2  1 which

 
implies that p  2 mod 7 and p  1  3 mod 4 . Applying (7.17) we have  
 7  p  2     
             1   2  (‡)
 p    
 7  Because p2 mod 7  7   7   7 
By (7.17)

2
From the previous case we have    1 and from the penultimate case we have
 7 
 1
   1 . Putting these into the above calculation (‡) yields
 7 
 Complete Solutions 7.4 Page 23 of 25

 7     
    1   2    1  1  1 .
 p   7   7 
Hence, we have proven that 7 is a quadratic residue of the primes p which satisfy
the congruence:


p  1,  3,  9 mod 28 . 
■
(a) We are asked to find the prime factorization of 120  7  14 393 . Using the
2

result of the above theory we have the prime factor must be of the form

 
p  1,  3,  9 mod 28 . The first few primes of this format are 3, 19, 29,
31, 37 and 37 is a factor of 14 393 because 14 393  37  389 . Also 389 is prime
because if it has a prime factor it would be less than 19 and the only factors below
19 of the given format is 3 and 19 and none of these are factors of 389 because
they were not factors of 14 393. Hence 14 393  37  389 .
(b) Similarly, we have to factorize 3542  7  125 309 . Let p be a factor of this

 
number. Then p  1,  3,  9 mod 28 . The first few are 3, 19, 29, 31, 37, --- .
By trialling these primes we find that 125 309  29  4321 . We need to find the

prime factors of 4321. First  4321   65 . There is no point trailing 3 and 19 as

 
these not factors of 125 309 so they cannot be factors of 4321. The next prime is
29 and we have 4321  29  149 and 149 is prime. Therefore, the prime
factorization of 3542  7  125 309 is 125 309  292  149

20. We need to show that one of the prime factors of x 2  3 is of the form 12n  7 .
Proof.
Let p  3 be a prime factor of x 2  3 . We have


x 2  3  0 mod p  
implies x 2  3 mod p . 
 
We have a quadratic congruence x 2  3 mod p . We need to show that 3 is a
quadratic residue for a prime p of the form 12n  7 . This implies that


p  7 mod 12 . 
Using Legendre symbols we have
 3  1  3 
        (†)
 p   p   p 
By (7.11)
 Complete Solutions 7.4 Page 24 of 25

 1 
    
 1 if p  1 mod 4
 
 p  1 if p  3 mod 4


  
 
From this p  7 mod 12 we have p  12k  7  4 3k  1  3 . Therefore


p  3 mod 4 
By (7.11)
 1
   1
 p 
By result of question 11 of Exercise 7.3 we have

 3  
   
 1 if p  1 or 11 mod 12
  
 p  1 if p  5 or 7 mod 12



 
 3 

   1 because p  7 mod 12 .
 p  
 1  3
Putting these two results    1 and    1 into (†) gives
 p   p 
 3   1  3 
         1  1  1
 p   p   p 


Hence 3 is a quadratic residue of a prime p  7 mod 12 which implies it is of 
the form 12n  7 . As we have a solution to the quadratic congruence

 
x 2  3  0 mod p so a prime factor of x 2  3 is of the form 12n  7 . This
completes our proof.
■

21. We are asked to prove that there are infinite number of primes of the form
3n  1 .
Proof.
Suppose there are a finite number of primes of the form 3n  1 and they are all
p1, p2 , , pn (*)

Consider the number N  3  p1  p2    pn   3 . Clearly 3 is a factor of N. Let

2

p  3 be another prime factor of N and x  3  p1  p2   pn then


N  x 2  3  0 mod p . 
 Complete Solutions 7.4 Page 25 of 25

 
This quadratic congruence x 2  3 mod p has solutions so 3 is a quadratic residue
of p. By question 11(ii) of Exercise 7.1 we have


3 is a QR of p  p  1, 11 mod 12 . 
   
This implies that p  1 or 11 mod 12 . If p  1 mod 12 then p  12k  1 but if
all the prime factors of N are of this form p  12k  1 then N must also be of this
form (you can show by induction that this is indeed the case) but it is not because

N  3  p1  p2    pn   3 .
2

 
So one of the factors must be of the form p  11 mod 12 which implies

 
p  12k  11  3 4 k  1  1 .

Hence p is of the form 3n  1 . Since p is a prime factor of N so p N . In view of p

being of the form 3n  1 it must be one of the primes in the above list (*). So
p x implies p x 2 .

Since p N and p x 2 so from N  3  p1  p2    pn   3 we must have

2

p 3.
This is impossible because p  3 . We have a contradiction so there are an infinite
number of primes of the form 3n  1 .
■