620-113 Explorations in Number Theory

Theorem 1 (Euclid, Prop. 30, Book VII). If p is prime and p | ab

then either p | a or p | b.
Proof. We have two choices - either p | a or p - a. If the former is true
then we have succeeded. However if p - a then there exists an integer c
such that pc = ab. Since p - a then p and a are coprime. By Theorem
5 there exists integers x and y such that 1 = px + ay. Therefore
b = b(px + ay)
= pbx + aby
= pbx + pcy
= p(bx + cy)
and thus p | b.

Remark 1. This only works for primes - e.g. 6 | 12 = 3.4 but 6 - 3 and
6 - 4.
Example 1. Factor the following integers into prime factors
(1) 540 = 22 .33 .5
(2) 3750 = 2.3.54
(3) 4725 = 33 .52 .7
(4) 3718 = 2.11.132
(5) 3234 = 2.3.72 .11
Theorem 2 (The Fundamental Theorem of Arithmetic). Except for
the order of the factors every positive integer n > 1 can be expressed
uniquely as a product of primes
r
Y
1 2
r
(1)
n = p1 p2 pr =
pi i
i=1

Proof. Proof by contradiction. Let

n = p 1 p 2 p r = q1 q 2 qs

r, s > 1

and suppose we order the prime factors

p1 p2 pr

and q1 q2 qs

and without loss of generality r s. The it is immediate that p |

q1 q2 qs and from Theorem 1 we know that p1 | qi for some 1 i s.
But p1 and qi are both primes so this implies p1 = qi , and furthermore
that p1 q1 . A similar argument can be applied for q1 , namely that
q1 | p1 p2 pr and by Theorem 1 q1 | pj for some 1 j r. This
1

implies q1 = pj p1 . Combining both these results we have in fact

p1 = q1 . Now divide out these equal factors to yield
n
= p 2 p r = q2 qs .
p1
The above argument can be repeated to show that
p 2 = q2
..
.
p r = qr .
If r < s then we would have
n
= 1 = qr+1 qs ,
p1 p2 pr
which is false because each qj > 1. Therefore we have r = s and pi = qi
for 1 i r and the representation is unique.

Theorem 3 (Euclid, Prop. 20, Book IX of the Elements). The number
of primes is infinite.
Euclids proof. Proof by contradiction. Suppose the number of primes
is infinite and these are p1 , p2 , . . . , pr . Consider the integer n = p1 p2 . . . pr +
1. Clearly n > pi for all i = 1, . . . , r and therefore must be composite. By Theorem 1 of the first Lecture n must have prime divisors pi .
Therefore pi | n for some i, but this implies pi | 1 which is impossible.

Kummers proof. Suppose there are only a finite number of primes p1 <
p2 < . . . < pr and let N = p1 p2 . . . pr > 2. Now consider N 1, which
being a product of primes has a prime divisor pi . Therefore pi | N 1
and pi | N which implies pi | N (N 1) = 1 which is impossible.