INTRODUCTION

Number theory is the branch of Mathematics, which deals with the properties
of integer, more specifically the properties of positive integers. Number theory is one
of the oldest branches of pure Mathematics and one of the largest courses; it contains
questions about number, usually meaning whole numbers or rational numbers.

Continued fractions are important in many branches of Mathematics. They

arise naturally in long division and in the theory of approximation to real numbers by
rational. These objects that are related to number theory help us to find good
approximations for real life constants.

The subject of continued fractions is an old subject although many people are
not aware of it. Actually, continued fractions have so many applications in algebra
and in various fields such as Mathematics, physics and chemistry.

The easiest way of forming a continued fraction is by writing a certain amount

in the form of a numerator and a denominator, and each denominator is composed of a
numerator and a denominator and a denominator and so on. Usually, the successive
numerators are equal to one.

Continued fractions have a long history; they were known since the appearance
of Euclidean algorithm for finding the greatest common divisor (GCD) of two
numbers. That was around the year 300 B.C.

Continued fractions is a different way of looking at numbers. It is one of the

most powerful and revealing representation of numbers that is ignored in Mathematics
that we’ve learnt during our study stage. A continued fraction if a way of representing
any real number by a finite (or infinite) sum of successive divisions of numbers.

The use of continued fraction is also important in mathematical treatment to

problem arising in certain applications. Such as calendar construction, astronomy,
music and others.

This project I have described about “continued fractions”. It contains some

sample of different theorem.

1
This project contains of Four chapters.

In chapter I, we discuss about some basic definitions in continued fractions.

In chapter II, we study about properties of continued fractions.

In chapter III, we shall explore some periodic of continued fractions.

In chapter IV, we deal with some application continued fractions.

2
 CHAPTER-I

BASIC DEFINITIONS

Definition: 1.1 [Continued Fraction]

A continued fraction is an expression in the form

b1
a 0+
b2
a 1+
b3
a 2+
b4
a 3+
b5
a4+
b
a5 + 6
…

In general, the numbers a 0 , a 1 ,a2, a3, a4 …, b1, b2, b3, b4 …. may be any real or
complex numbers or functions of such variables. The number of terms can be finite or
infinite.

Definition: 1:2 [Simple Continued Fraction]

A simple continued fraction is an expression of the form

1
a 0+
1
a 1+
1
a 2+
1
a 3+
1
a4+
1
a5 +
…

Where a0 is an integer and the terms a1, a2, a3, a4, … are positive integers.

Definition: 1.3 [Finite Simple Continued Fraction]

3
 A finite simple continued fraction has only a finite number of terms, with the
form.

1
a 0+
1
a 1+
1
a 2+
1
a 3+
1
…+
1
a n−1+
an

Where a 0 is an integer and the terms a 1 , a2 , a3 , … an are positive integers.

Definition: 1.4 [Rational Number]

A rational number is a number that can be in the form of p/q. where p and q are
integers and q is not equal to zero.

Definition: 1.5 [Integer]

The numbers … -4, -3, -2, -1, 0, 1, 2, 3, 4, … are called integers. We

represented the set of integers by z.

Thus Z={ …−4 ,−3 ,−2,−1 ,0 ,1 , 2 ,3 , 4 , … }

Definition: 1.6 [ Purely Periodic Continued Fraction]

A purely periodic continued fraction with period (or length) m is a continued

fraction such that the initial block of partial quotients a 0 , a 1 , a2 , a3 , … am −1 is repeated
over and over (so that a m=¿a , a ,a , a ,… a
0 1 2 3 2m−1=¿am −1 ¿ ¿ and so on for each a kmWhere k≥ 1) and no
shorter block a 0 , … a n−1 with n¿ m has this property.

A purely periodic continued fraction will be represented as [ a 0 , a 1 , a 2 , a3 , … am −1 ]

4
Definition: 1.7 [ Periodic Continued Fraction]

A periodic continued fraction consists of an initial block of length n followed by a

repeating block of length m. It is supposed that there is no shorter repeating block and
the initial block does not end with a copy of the repeating block.

A periodic continued fraction will be represented as

[ a 0 ,a 1 , a 2 , a3 , … a n−1 , a ,a , a ,a , …a ]
0 1 2 3 m −1

Definition: 1.8 [Quadratic Surd]

A quadratic surd is a solution of a quadratic equation ax2+bx+c=0 with integer

coefficients a, b and c. where a ≠ 0 and its discriminant b2-4ac is not a perfect square.

Definition: 1.9 [Complete Quotient]

We call a m=[ a m , am+1 , … an ] the mth complete quotient of the continued fraction
'

[ a 0 , a 1 , a 2 , a3 , … an ]

Let x¿ [ a 0 , a 1 , a2 , a3 , … an ]. Then

' '
' a1 a0 +1 an p n−1+ p n−2
x¿ a 0= '
= '
a1 a n q n−1+ qn−2

Definition: 1.10 [Convergent]

We call [ a 0 , a 1 , a 2 , a3 , … am ] (for 0 ≤ m≤ n ¿ the mth convergent to

[ a 0 , a 1 , a 2 , a3 , … an ].

In our example, the convergent are

2
1) 2=
1
1 7
2) 2+ =
3 3

5
 1 9
2+ =
3) 1 4
3+
1
1 43
2+ =
1 19
4) 3+
1
1+
4

Definition: 1:11 [Congruence]

If an integer m ≠0 divides the difference a-b, then it is said ‘a’ is congruent to

‘b’ modulo m and is written as

a ≡ b(modm)

For example

1) 30≡ 0(mod 5)
2) 67≡2 (mod 13)

Definition 1:12 [Quadratic Residues]

If m is a positive integer, we say that the integer a is a quadratic residue of m if

(a, m) = 1 and the congruence x2≡ a(mod m) has a solution.

If the congruence x2≡ a(mod m) has a no solution, we say a is quadratic

nonresidue of m.

6
 CHAPTER-II

PROPERTIES OF CONTINUED FRACTIONS

2.1 FINITE CONTINUED FRACTIONS

2.1.1 Rational Numbers

Theorem 2.1

Any rational number p/q can be represented as a finite simple continued

fraction in which the last term can be modified so as to make the number of terms in
the expansion either even or odd.

Proof

Suppose that

p
=[a0 , a1 , a2 , a3 , … an ]
q

If a n=1, then

1 1
=
1 an−1+ 1
an−1 +
an

and it is possible to represent

p
=[ a0 , a1 , a2 , a3 , … an−2 , an−1+ 1 ]
q

If a n> 1, then

1 1
=
an 1
(a¿¿ n−1)+ ¿
1

7
and it is possible to represent

p
=[ a0 , a1 , a2 , a3 , … an−2 , an−1−1 ,1 ]
q

To avoid this ambiguity, the last partial quotient a n of any finite continued
fraction will be restricted to be greater than 1 from now on.

Now, consider the following:

35 35 1 1 1 1
=0+ =0+ =0+ =0+ =0+
93 93 93 23 1 1
2+ 2+ 2+
35 35 35 12
1+
23 23

1 1 1
¿ 0+ =0+ =0+
1 1 1
2+ 2+ 2+
1 1 1
1+ 1+ 1+
23 11 1
1+ 1+
12 12 12
11

1
¿ 0+ =[ 0 , 2, 1 ,1 , 1 ,11 ]
1
2+
1
1+
1
1+
1
1+
11

93
Compare this with =[ 2 ,1 , 1 ,1 , 11] .
35

In fact, it can be generalized to the following theorem.

Theorem: 2.2

8
 p
For integers p, q where p¿ q , = [ a0 , a1 , a2 , a3 , … an ] if and only if
q
q
=[ 0 , a0 , a1 , a2 , a3 ,… an ].
p

Proof

p
If p¿ q >0, then >1 an
q

p
=[ a 0 , a 1 , a 2 , a3 , … an ]
q

1
¿ a 0+
1
a 1+
1
a 2+
1
a 3+
1
a4+
1
...+
an

p
Where a 0 is an integer ¿ 0. The reciprocal of
q
is

q 1 1
= =
p p 1
a 0+
q 1
a 1+
1
a 2+
1
a 3+
1
a4+
1
…+
an

1
¿ 0+
1
a 0+
1
a 1+
1
a 2+
1
a 3+
1
a4+
1
…+
an

¿ [ 0 , a0 , a1 , a2 , a3 , … an ]

9
 q
Conversely, if q < p then is of the form
p

q 1
=0+
p 1
a0 +
1
a1 +
1
a2 +
1
a3 +
1
a4 +
1
…+
an

And its reciprocal is

p 1
=
q 1
1
a0 +
1
a 1+
1
a2 +
1
a3 +
1
a4+
1
...+
an

1
¿ a 0+
1
a 1+
1
a 2+
1
a 3+
1
a4+
1
...+
an

This concludes the discussion of representing rational numbers by continued fractions.

2.1.2 Inverting a Fraction

Given a non-zero rational number, we simply interchange the numerator and

denominator to get its reciprocal.

43 19
For example, the reciprocal of is .
19 43

10
 Now we describe how to find the reciprocal of a rational number if it is
described as a simple continued fraction:

1. If the simple continued fraction has a 0 as its first number, then remove the
0.

2. If the simple continued fraction does not have 0 as its first number, then shift
all the numbers to the right and place 0 as the first entry.

Examples:

43 19
=[ 2 , 3 ,1 , 4 ] ⇒ =[ 0 ,2 , 3 ,1 , 4 ]
19 43

3 7
=[ 0 , 2, 3 ] ⇒ =[ 2 ,3 ]
7 3

2.2 MULTIPLE CONTINUED FRACTIONS

Given a rational number, we have seen one way of constructing a simple

continued fraction (namely by Euclid’s algorithm). In this section we will see that
there is essentially unique way to write a rational number as a simple continued
fraction.

Theorem 2.3

If x is representable by a simple continued fraction with and odd (even) number

of convergents, it is also representable by one with an even (odd) number.

Proof

If a n ≥ 2,

[ a 0 , a 1 , a 2 , a3 , … an ]=[ a0 , a1 , a2 , a3 , … an−1 , 1 ]

If a n=1,

11
 [ a 0 , a 1 , a 2 , a3 , … an −1 , 1 ]=[ a 0 , a 1 , a2 , a3 , … an−1 +1 ],[ 1 ] =[ 0 , 1 ]

Thus, the proof of this theorem says that there are atleast two ways of writing a
simple continued fraction for a rational number.

1. A simple continued fraction ending with some m¿ 1 i.e. [ … , m ].

2. A simple continued fraction ending with 1 i.e. replace the final m by (m −1 ¿
+1/1 to get [ … m−1 , 1 ].

Examples:

[ 1 ,2 , 3 , 4 , 5 ] = [ 1 , 2 ,3 , 4 , 4 , 1 ]

3 1 1
=1+ =1+
2 2 1
1+
1

2.3 RELATIONS BETWEEN CONVERGENTS

In this section, we see some properties of the simple continued fractions in

terms of the numerators and denominators appearing in the convergents.

Theorem: 2.4

If pn and q nare defined by p0=a0, p1=a1 a0+1, pn=¿¿ a n pn−1 + pn−2 for 2 ≤ n

q 0=1 , q1=a1 , qn=an qn−1 +q n−2 for 2≤ n,

then

p
[ a 0 , a 1 , a 2 , a3 , … an ]= q n .
n

Proof

The proof proceeds by induction. The base cases are seen to be true by the
assumptions given for n=0 , n=1. Let us assume the statement to be true for some m.
Then

12
 p a p +p
[ a 0 , a 1 , a 2 , a3 , … am −1 , am ]= q m = am qm −1 +q m −2
m m m −1 m−2

Hence, we get

[ a 0 , a 1 , a 2 , a3 , … am −1 , am , a m+1 ]= [ a0 , a1 , a2 ,a 3 , … a m−1 ,a m ,
1
a m+1 ]
(¿ a + a1 ) p
m
m +1
m −1 + pm −2

( a + a1 ) q
m
m +1
m −1 +q m−2

am +1 ( am p m−1+ p m−2 ) + p m−1

¿
am +1 ( am qm −1 + qm−2 ) +q m−1

p m+1
¿
q m +1

By the principle of mathematical induction, pnand q n are indeed defined by the

recursive relation stated in the theorem.

It follows that the nth convergent is

p n an p n−1 + pn−2
=
qn an q n−1 +q n−2

Theorem 2.5

The numbers pn and q n satisfy pn q n−1 −p n−1 q n=(−1 )n−1

Proof

From the previous theorem, we have

pn q n−1 −p n−1 q n=( an p n−1 + pn −2 ) q n−1− pn−1 ( an qn −1 +q n−2 )

¿−( pn−1 qn−2− pn −2 q n−1)

13
Repeating this step with n−1 ,n−2 , … ,2 in place of n, gives us

pn q n−1 −p n−1 q n=(−1 )n−1 ( p 1 q 0− p 0 q 1 ) ¿ (−1 )n−1 (1 )=(−1 )n−1

Example:

225 1
=1+
157 1
2+
1
3+
1
4+
5

Its convergents are

1
1=
1

1 3
1+ =
2 2

1 10
1+ =
1 7
2+
3

1 43
1+ =
1 30
2+
1
3+
4

1 225
1+ =
1 157
2+
1
3+
1
4+
5

1 3 10 43 225
i.e. , , , ,
1 2 7 30 157

The numerators and denominators of these convergents satisfy

( 3 ) ( 1 )−( 1 ) ( 2 )=(−1 )1−1=1

14
 ( 10 ) ( 2 )−( 3 )( 7 )=(−1 )2−1=−1

( 43 )( 7 )−( 10 ) ( 30 )= (−1 )3−1=1

( 225 ) ( 30 )− ( 43 ) ( 157 ) =(−1 )4−1=−1

Theorem 2.7

a m=[ a'm ], the integral part of a 'm, except that a n−1=[ an−1 ]−1 when a n=1.

Proof

If n=0, then a 0 ¿ a '0=[ a'0 ]. If n> 0, then

' 1
a m=a m + ' for ( 0 ≤ m≤ n−1 )
a
m+1

Now

'
a m+1 >1 for ( 0 ≤ m≤ n−1 )

except that a 'm+1=1 when m=n−1 and a n=1.

This is because a 1 , a2 , a3 , … , an are all non-negative integers and inductively one

can see that the above statement is true.

Hence

'
a m <am < am +1 for ( 0 ≤ m≤ n−1 )

and

a m=[ a'm ] for ( 0 ≤ m≤ n−1 )

except in the case specified. And in any case

a n=a'n=[ a 'n ]

2.4 UNIQUENESS CONTINUED FRACTIONS

15
 In this section we use all the properties seen in the above theorems to show that
under some minor conditions, every rational number has a unique finite simple
continued fraction.

Theorem: 2.8

If two simple continued fractions [ a 0 , a 1 , a 2 , a3 , … , an ] , [ b 0 , b 1 , b , b3 , … , b N ] have

the same value x, and a n> 1, b N > 1, then n=N and the fractions are identical.

Proof

By theorem 2.7, a 0=b 0=integral part of x . Let us assume that the first m terms
in the continued fractions are identical. Then

x=[ a0 , a1 , a2 , a 3 , … , a m−1 , a'm ]=[ b 0 ,b 1 , b 2 , b3 , … , bm −1 , b'm ]

If m=1, then

1 1
a 0+ '
=b0 + '
a1 b1

which implies a '1=b'1 and by Theorem 2.7, a 1=b1. If m>1, then

' '
am pm−1 + pm−2 bm p m−1+ p m−2
'
= '
a m q m−1 +q m−2 bm q m−1+ qm −2

( a'm−b'm ) ( pm −1 q m−2− pm−2 qm −1 ) =0. But ( pm −1 q m−2− pm−2 qm−1 ) =(−1 )m, by Theorem 2.5 and
so a 'm=b 'm . By Theorem 2.7, a m=b m.
Suppose now, n ≤ N , then we have shown that a m=b m ∀ m ≤n . If N >n , then
pn
=[ a0 , a1 , a2 , a3 , … an ]=[ a0 , a1 ,a 2 , a 3 , … a n , b n+1 ,b n+2 , … , b N ]
qn
'
bn +1 pn + pn−1
¿ '
b n+1 qn +q n−1
⇒ pn q n−1− pn−1 qn=0
which contradicts theorem 2.5. Hence n=N and the fractions are identical.

16
 Theorem 2.3 and Theorem 2.8 together tell us that there are exactly two ways
of writing any rational number as a finite simple continued fraction. They also tell us
how to convert one simple continued fraction to the other. Since we already know
how to obtain a simple continued fraction by using the Euclid’s algorithm, this is
essentially the only way to obtain a simple continued fraction.
Example:
[ 7 ]= [6 , 1 ]
[ 1 ,2 , 2 ,2 ] =[ 1 , 2 ,2 , 1 ,1 ]
[ 0 , 1, 2 , 3 ] =[ 0 , 1, 2 , 2 ,1 ]
[ 1 ,1 , 1 ,1 , 1 ] =[ 1, 1 ,1 , 2 ]
[ 1 ] =[ 0 , 1 ]

17
 CHAPTER-III

PERIODIC CONTINUED FRACTIONS

A periodic continued fraction will be represented as

[ a 0 , a 1 , a 2 , a3 , … . an−1 , an , an+1 , … , an +m−1 ]

Theorem: 3.1 (Euler)

If t is periodic continued fraction, then t is a quadratic surd.

Proof

Case 1: Assume t is purely periodic and has the form [ a 0 ,a 1 , … , a m−1 ]. Then

t=r 0 =r m=r 2m =… by the statement

pk −1 r k + pk−2
“For any k ≥ 1 , [ a 0 , a 1 , a 2 , a3 , … an ]= ”
qk −1 r k +q k−2

p r +p p t+p
[ a 0 , a 1 , a 2 , a3 , … ]= qm−1 r m +q m−2 = q m−1 t +q m−2
m−1 m m−2 m−1 m−2

and so

q m−1t2 +( q m−2− pm −1 ) t− pm−2=0

18
Since m>0, q m−1 ≥ q 0=0 and hence t is a quadratic surd.

Case 2: If t has initial terms and has the form [ a 0 , a 1 , a 2 , a3 , … . an−1 , an , an+1 , … , an +m−1 ],
then r n =r n+m =r n +2m =… and using twice by the statement “For any k ≥ 1 ,
p r +p
[ a 0 , a 1 , a 2 , a3 , … an ]= qk −1 r k +q k−2 ’’
k −1 k k−2

p n−1 r n + pn −2
t=
q n−1 r k + qn−2

And

p n+m −1 r n + pn+ m−2

t=
q n+m−1 r k +q n+m −2

Then solve for r n :

−qn−2 t− p n−2
rn =
qn−1 t− p n−1

and

−qn+ m−2 t− pn +m−2

rn =
qn+ m−1 t− pn +m−1

After some algebraic manipulation, it can be deduced that

( q n−2 qn+ m−1−q n−1 q n+m −2 ) t 2+ ( −q n−2 pn+ m−1−p n−2 q n+ m−1+ p n−1 q n+m−2 +q n−1 pn +m−2 ) t+ ( pn −2 pn+ m−1−p n−1 pn +m−2 )

If this is not a quadratic equation, then q n−2 q n+ m−1 will be equal to q n−1 q n+m −2 and
then q n+m −1 will divide q n−1 q n+m −2. As q n+m −1and q n+m −2 are relatively prime, q n+m −1 will
have to divide q n−1 and this is not possible as q n+m −1 >q n−1. Contradiction. Thus, this is
a quadratic equation and t is a quadratic surd.

Theorem 3.2 (Lagrange)

If t is a quadratic surd, then t is a periodic continued fraction.

Proof

Since t is a quadratic surd, there are integers d 0¿ 0 , b0 and c 0 such that

2
d 0 t +b 0 t +c 0=0. By using the theorem

19
 p r +p
“For any k ≥ 1 , [ a 0 , a 1 , a 2 , a3 , … an ]= qk −1 r k +q k−2 ” to replace t with an
k −1 k k−2

expression in an arbitrary r k +1,

2 2
d 0 ( pk r k+ 1+ p k−1 ) +b 0 ( p k r k−1+ p k−1) ( qk r k+ 1+ qk−1 ) +c 0 ( q k r k+1 +q k−1 ) =0

After some algebraic manipulations. This becomes

2
d k +1 r k+1 +b k+1 r k+1 +c k +1=0
where
2 2
d k +1=d 0 p k +b0 pk q k + c 0 q k

b k+1=2 d 0 p k p k−1+ b0 ( p k qk−1 + pk−1 q k ) + 2 c0 q k qk−1

2 2
c k +1=d 0 p k−1+ b0 p k−1 q k−1 +c 0 qk−1

Such that d k =c k+1. If d k +1=0 , this would not be a quadratic equation and r k +1 would be
a rational number and then so would t, hence d k +1 ≠ 0 for all k ≥ 0. The discriminant is
2 2 2
b k+1−4 d k +1 c k +1=(b0−4 d 0 c 0 ) ( pk q k−1+ p k−1 q k )

and this remains unchanged.

1 1 ϵ
As |q k t−p k|< q <
qk , it can be written that pk =qk t+
q k where
k+1

|e|<1. Then

( ) ( )
2
∈ ∈ 2
d k +1=d 0 qk t+ +b0 qk t+ qk+ c0 qk
qk qk

( )
2
∈
¿ ( d 0 t + b0 t+ c 0) p + ∈ ( 2 d 0 t+b 0 ) +d 0
2 2
k
qk

( )
2
∈
¿ ∈ ( 2 d 0 t+ b0 ) + d 0
qk

and note that

|d k+1|<|2 d 0 t|+|d 0|+|b0|

Which means |d k+1|is bounded. As d k =c k+1, |ck +1| is also bounded and same for
b k+1, since the discriminant is a constant. Thus the coefficients of all the equations are

20
bounded and there can be only finitely many such equations. Therefore, the equations
will have to repeat and the corresponding r k +1 ' s will coincide. Thus t is a periodic
continued fraction and the proof is complete.

Now the period length of a periodic continued fraction will be investigated. By

( p0 + √ D)
quadratic formula, either solution to ax2+bx+c=0 can be written as t= where
Q0

Q0 ≠ 0. p0 and D>0 are integers such that D is not a perfect square and Q 0|( D− p20 ) . To

develop t as a continued fraction, set a 0=[t ] and then

1 Q0 p +√ D
r 1= = = 1
t−a 0 p0 −a0 Q0 + √ D Q1

where

p1=a0 Q 0− p0

and
2 2
D− p0 +2 a 0 p0 Q0−( a0 Q0 )
Q 1=
Q0

2
D− p1
are integers. As =Q0, p1 and Q1 have the same divisibility property for p0 and
Q1
Q 0. Continue this process, it can be obtained that

p k+1 + √ D
r k +1=
Qk+1

2
(D− p k+1 )
where pk +1=a k Qk −p k and Qk +1= is not zero since D is not a perfect
Qk

square. So D− p 2k+1=Q k+1 Q k and for sufficiently large k’s such that

0< P k+1 < √ D

And

0< √ D−Pk+1 <Qk +1< √ D + Pk +1< 2 √ D

21
 As the inequalities 0< P k+1 < √ D and 0<Q k+1 <2 √ D make sure that there can be
no more than √ D× 2 √ D distinct pairs { Pk +1 ,Qk +1 }, the period length L of any surd of
discriminant D satisfies L ( D )< 2 D . This is the Lagrange’s estimate. The following
theorem is a more precise statement.

Theorem 3.3

( p ¿ ¿ 0+ √ D)
Let t= ¿ be any quadratic surd where Q0 ≠ 0. p0 and D>0 are
Q0
integers such that D is not a perfect square and Q0 ¿. Then the length L of the repeating
block in the periodic continued fraction of t satisfies.

L ( D )=O( √ D log ( D )) ,

Where O() is the Big Oh asymptotic notation.

Proof

From above, it suffices to estimate the number of pairs { P , Q } of integers

satisfying

0< P< √ D

0< √ D−P<Q< √ D+ P<2 √ D

and

P2 ≡ D mod Q

If Q> √ D , then √ D<Q< √ D+ P<2 √ D and hence

0<Q−√ D< P< √ D ,

while if Q< √ D , 0< √ D−P<Q< √ D gives

0< √ D−Q< P< √ D

In either case, given a value of Q, the possible values for p are contained in an
interval of less than Q, and the number of possible p’s can be counted by their residue
classes modulo Q. Thus

22
 L ( D )< ∑ ( ∑ 1
)
0<Q <2 √ D P 2 ≡ D modQ

Consider the inner sum

∑ 1
P 2≡ D mod Q

for the 3 possible cases (Q, D) =1,1 < (Q, D) < Q and (Q, D) = Q. In the first case,
assume P2 ≡ D mod Q is solvable, that is, D is a quadratic residue of Q and let prime
factorization of Q be
k0 k1 km
Q=2 P1 , … , P m

Where P1…Pm are distinct odd primes with positive integer exponents k 1…km
and the integer exponent of 2 is k0 ≥ 0. Then any solution of P2 ≡ D mod Q must also
satisfy the congruences P2 ≡ D mod 2 k , P2 ≡ D mod P1k , …, P2 ≡ D mod Pmk
0 1 m

If the number of solutions of N i to each of these congruences (i = 0, 1, 2, 3, …,

m) is known, by the Chinese Reminder Theorem, there are exactly N 0N1…Nm
solutions to the original congruence.

For any of the odd primes pi, a solution of P2 ≡ D mod Pik means there is a i

solution of P2 ≡ D mod Pi, and there are exactly two such solutions. By Euler’s lemma,
( p−1)
if there exists a number k such that k 2≡ a (mod p), then a 2
≡ 1(modp). In Legendre
symbol, if p is an odd prime, then for every c,
( p −1 )

()c
p
≡c 2
( mod p ) .

Thus the solutions pull back to two solutions modulo Pki and so Ni = 2 for i = 1, i

2, …, m. For the prime 2, the situation is slightly different. Any odd number a is
always a square modulo 2, and a is square modulo 4 only if a ≡ 1(mod 4). And a is a
square modulo 2z only if a ≡ 1(mod 8) for z ≥ 3.Thus given a solution of p2≡ D mod2,
there will be two solutions modulo 4 and four solutions modulo 2 z for z ≥ 3 . For the
case (Q, D) =1,

2
∑ 1=2 2 ,
k m

p ≡ D modQ

23
where k= 0 if k0 ≤0, k = 1 if k0 =2 and k=2 if k0 ≥ 3, and m is the number of distinct
odd prime divisors of Q. Since 2k+m ≤ τ (Q), where τ (n) is the number of divisors of the
integer n, it is shown that

2
∑ 1 ≤ τ (Q)
p ≡ D modQ

The contribution to this sum for a Q in the second case, when 1 < (Q, D) < Q,
can be no more than that from the first case since for these Q’s the congruence
p ≡ D modQ , is the same as
2

( ) ( )
2
(Q , D) p D Q
≡ mod
(Q , D) (Q , D) (Q , D)

Which either has no solution or at most τ ( (QQ, D) ) solutions, as in the first case.
Finally if Q divides D.

2
∑ 1=O( √ D).
p ≡ D modQ

Since

[]
n n
n
∑ τ (k )=∑ d
=nlog (n)+O(n).
k =1 d=1

the estimate that

L=O
( ∑ τ (Q)
)
o <Q <2 √ D

is equivalent to

L(D)=O( √ D log (D)).

Hence the theorem is established.

24
 CHAPTER-IV

APPLICATIONS

4.1 CALENDAR CONSTRUCTION

There are many activities whose success depends on accurate planning. Some
of them should be done during a certain period of a year, such as sowing and plowing.
So, calendar construction is an important issue since ancient times and calendars are
found in every old civilization.

By counting the days, calendars help us to determine the seasons which depend
on the rotation of the Earth around the sun.

A tropical “solar” year is the time it takes the Earth to make one revolution
315569259747
around the sun¿
864000000
= 365.24219878125 days.

Remark 4.1:

There are 365 days, 5 hours, 48 minutes and 45.9747 seconds in a year. So,
there are (365∗24+5+ 48 /60)∗3600+45.9747=31556925.9747 seconds in one year. On
the other hand, there are 24∗3600=86400 seconds a day. Babylonian Calendar “the
oldest” contained 12 months with 29 and 30 days alternately. One year in this
calendar had 354 days. After that, the Babylonian calendar was replaced by the
Egyptian Calendar which consisted of 12 months, each month contained 30 days. One
year in this calendar consisted of 360 days. Then, five days were added to adjust the
calendar in Pharaonic times. This calendar was effective for more than 3000 years.
However, it led to an error of quick accumulation and therefore a noticeable shift of

25
seasons. Next, a sixth day was introduced every fourth year to give a calendar called
the Alexandrian Calendar.

Our calendar comes from the ancient Roman calendar. A year in Roman
calendar consisted of 365 days until an Alexandrian astronomer advised Romans to
create the Julian calendar in which every year divisible by 4 was a leap year
“consisting of 366 days” and every other year was a common year “consisting of 365
days”. Julian calendar was a good calendar as it accumulated a small error in a
hundred years. However, over the next millennium, the discrepancy was noticed.

Finally, a new more precise calendar construction was created by pope Gregory
XIII. He decreed to omit a leap year every century except those years that are
divisible by 400. The Gregorian calendar is both accurate and easy to remember.

The question now is what is the science behind this construction? In fact, continued
fractions provide such a science.

The idea of constructing a modern calendar is to have a cycle of q years such

that p of them are leap years. So, q-p are common year. When p and q are chosen, we
should take into consideration that the mean year length is very close to the tropical

26
year. Moreover, the rule for selecting p leap years should be convenient and simple to
use.

During the q-cycle with p leap years, there are 365q+p days. Thus, the mean
p
year length is 365+ .
q

Recall that a tropical year consists of

315569259747 209259747 7750361

=365+ =365+ =365.24219878125
864000000 864000000 32000000

days.

p 7750361
Now, our purpose is to find a good approximation
q
for
32000000
. The last fraction
represents the error between a tropical year and a common year.

Representing this fraction as a continued fraction yields

7750361
=[0 , 4 ,7 ,1 , 3 ,5 , 6 , 1 ,1 , 3 ,1 , 7 , 7 , 1, 1 ,1 , 1 ,2 ]
32000000

The first 8 corresponding convergents are

1
c 0=0 c 1= =0.25
4

7 8
c 2= =0,24137931034 c 3= =0.2424242424
29 33

31 163
c4= =0.2421875 c 5= =0.24219910846
128 673

1009 1172
c 6= =0.24219875180 c 7= =0.24219880141
4166 4839

The Julian Calendar is realized by the first convergent c 1 which gives a 4-year
cycle with one leap in it.

The annual error considered in Julian Calendar is

|14 − 32000000
7750361
|=0.00780121875
27
 Which means that the calendar accumulates about 8 extra days in 1000 years.
That is a bit less than a day in 100 years.

Looking at the denominators next convergents, we realize that the numbers 29,
33, 128, 673, … provide uncomfortable lengths of a cycle. For example, the fourth
31
convergent determines a 128-year cycle with 31 leap years in it. We can
128
construct a corresponding calendar in which there is a leap year every fourth year in
the cycle with the thirty-second leap year deleted and this construction gives an annual
error

|128
31
−
32000000|
7750361
=|−0.00001128125|=0.00001128125

Which means a loss of about one day every 100000 years. This construction is
more accurate than Julian Calendar but it is uncomfortable to use. So, no one used
this calendar.

Now, we try to find a cycle several centuries long with easy and simple
selection rule of leap years. Suppose that q=100 t , t is an integer lies between 1 and 9.
This assumption matches with the problem of approximating the fraction

7750361 7750361
100 × = .
32000000 320000

7750361
= [ 24 , 4 , 1 ,1 , 4 ,1 , 2 ,2 , 6 , 11, 2 ,1 , 1, 2 ]
320000

The first corresponding 6 conergents are

97
c 0=24 c 1= =24.25
4

121 218
c 2= =24.2 c3 = =24.22222222222
4 9

993 1211
c4= =24.21951219512 c 5= =24.22
41 50

28
 97
The first convergent c 1= gives a 400-cycle with 97 leap years in it. The
4
selection rule of leap years in the cycle is that every year divisible by 4 is a leap year
except 100th, 200th, 300th years. This calendar is called Gregorian Calendar which is
used nowadays in most countries.

The error results in a century from this calendar is

|974 − 7750361
320000 |
=0.030121875

That is an accumulation of about one extra day every 3320 years.

121
Another calendar could be constructed using the convergent c 2= which
5
gives a 500-year cycle with 121 leap years in it. The selection rule of leap years in
this cycle is that every year divisible by 4 is a leap year except 100 th, 200th, 300th, 400th
years.

The error results from this calendar in a century is

|1215 − 7750361
320000 |
=|−0.019878125|=0.019878125

Which implies that there is a loss of nearly a day every 5031 years.

Moreover, we can construct a calendar using a 900-year cycle with 218 leap
years in it. The selection rule of leap years in this calendar implies 7 exceptions to the
fourth year rule since (900 ÷ 4 )−7=218. This calendar is accurate since it accumulates
only one day in about 42660 years. However, it is more complicated than the previous
calendars and the 900-years cycle is long and therefore inconvenient. So, we reject
this calendar and prefer the simpler ones.

A small correction can be done in future to the Gregorian Calendar to make it

more accurate. Continued fractions give an easy method to carry out his correction.
The idea is to find a longer cycle length q consisting of a number of 400-year cycles.
Suppose q=400s, where s is the number of 400 year cycles in the new longer cycle.

29
 7750361 7750361
Represent 400 × = as a simple continued fraction to get
32000000 80000

7750361
=[96 , 1 ,7 , 3 , 2 ,1 , 25 , 2, 1 ,5 ,2]
80000

The first corresponding four convergents are

c 0=96 c 1=97

775 2422
c 2= c3=
8 25

775
The second convergent c 2= provides us with a 3200-year cycle with 775 leap
8
years in it.

The error of this construction is

|7758 − 7750361
80000 |
=|−0.0045125|=0.0045125

Which implies that there is a loss of about one day every 88643 years.

Remember that in Gregorian Calendar, there are 97 leap years in every 400-
year cycle. So, we get 97×8 = 776 leap years within every 3200 years. Therefore,
omitting one leap year every 3200 years will provide us with a modified Gregorian
Calendar which has nearly the same construction as Gregorian Calendar but is more
accurate.

4.2 ASTRONOMY

Dutch physicist and astronomer Christiaan Huygens designed an automatic

planetarium to show relative motions of the six known planets around the sun (and
separately, the Moon around the Earth). Each model planet was attached to a large
circular gear which was driven by a smaller gear mounted on a common drive shaft.
35
Using the best data available in 1680 and approximating the year as 365 days.
144
Huygens calculated the ratios of the periods of the planet’s orbits to the Earth’s year
and obtained the following values.

30
 Gear ratios of the solar systems used by Huygens

planet Period convergent Gear teeth

Mercury ☿ 25335 p 5 33 33 :137
=
105190 q5 137

64725 p5 8
Venus ♀ 105190 = 32:52
q5 13

Earth ♁ 1

♂ Mars 197836
p 5 79
105190 = 158:84
q5 42

p 3 83
1247057 =
q3 7
Jupiter ♃ 105190 166:137

p 1 59
=
3095277 q1 2
105190 118:4
Saturn ♄

After expanding these ratios as continued fraction, he chose gear pairs corresponding
to convergents such that the component number would be practical to construct as
gear teeth.

After Huygens died, two more planets Uranus and Neptune are discovered.

Planet Period Ratio convergent Gear teeth

Uranus ⛢ 30707.4896 30707489600 p 1 1177 1177:14
=
365256366 q1 14

31
 60223352800 p 2 1319
=
365256366 q2 8
Neptune ♆ 60223.3528 1319:8

32
Photograph of Huygen’s automatic planetarium

33
Sketch of Huygen’s automatic planetarium (front)

Sketch of Huygens automatic planetarium (back)

34
 Eris, Pluto, Ceres, Moon and Earth

Gear ratios of dwarf planets

Planet Period Ratio Convergent Gear teeth

Ceres 1679.819 1679819000 p 3 23 115:25
=
365256366 q3 5

90613305500 p 1 2977
Pluto ♇ 90613.3055 365256366 = 2977:12
q1 12

p 1 3343
20350000 =
Eris 203500 q1 6 3343:6
36525

There are also three dwarf planets in the Solar System. They are Ceres, Pluto, and
Eris.

35
Note that as the periods of Pluto and Eris is very large comparing to Earth, it is not
quite possible to provide reasonable gear ratios. The same holds for Uranus and
Neptune.

In this paper, the periods of exoplanet systems of Gliese 876 and Mu Arac and
satellites of Saturn and Uranus are also investigated.

Gliese 876 is a red dwarf star located approximately 15 light-years away in the
constellation Aquarius. Currently, it is known to host three extrasolar planets.

Gliese 876 planetary system

The period of GJ 876 c is set to 1 in order for comparison.

Gear ratios of the planetary system of Gliese 876

Extrasolar Orbital Ratio Convergent Gear teeth

planet period(days)
GJ 876 c 30.340 1 - -

6094 p 1 233
GJ 876 b 60.940 3034
= 233:116
q1 116

193776 p1 3
GJ 876 d 1.937760 3034000 = 9:141
q1 47

36
Mu Arae is a Sunlike yellow-orange star located around 50 light years away in the
constellation Ara. The star has a planetary system with four known planets. The
period of mu Ara e is set to 1.

Mu Ara e

Gear ratios of the planetary system of Mu Arae

Extrasolar Orbital period(days) ratio Convergent Gear teeth

planet

mu Ara e 310.55 1 - -

63000 p 2 71
mu Ara b 630.0 = 71:35
31055 q2 35

249000 p 1 441
mu Ara c 2490 31055 = 441:55
q1 55
955
p3 2
mu Ara d 9.550 31055 = 4:130
q3 65

37
Similarly, the same method can be applied to the moons of Saturn (♄). Saturn has 60
moons as of now (57 of them are confirmed and named)

The Saturnian system (photographic montage)

Here some of them are considered. The period of the largest moon. Titan, is set to 1
here for comparison.

Gear ratios of the Saturnian system

Moon Orbital Ratio Convergent Gear teeth

period(days)
VI Titan 15.945421 1 - -

9424218 p 3 12
I Mimas 0.9424218 = 12:203
159454210 q3 203

1370218 p 5 11
II Enceladus 1.370218 15945421 = 11:128
q5 128
79330183
p 2 199
VIII Iapetus 79.330183 15945421 = 199:40
q2 40
550564636
IX Phoebe 550.56436 15945421 p 2 69 138:4
=
q2 2
\

38
 Gear ratios of the Uranian system

Moon Orbital Ratio convergent Gear teeth

period(days)
I Ariel 2.520379 1 - -

4144177 p 6 171
=
II Umbriel 4.144177 2520379 q6 104 171:104

8705872 p 3 38
III Titania 8.705872 = 156:44
2520379 q3 11

Uranus (⛢) is the seventh planet from the Sun, and the third largest planet in size. It
has 27 known satellites. The period of Uranus I is set to 1 for comparison.

4.3 PIANO TUNING

Musicians know that we cannot tune a piano perfectly. In this discussion, we

study the role of continued fractions in piano tuning.

The keyboard of a piano consists of white and black keys. The standard white
keys are A, B, C, D, E, F and G. A black key is called sharp and flat. It is a sharp key
“#” of the white key that precedes it and a flat key “b” of the white key that follows it
as figure shows.

39
 Sounds that have frequencies with small integer ratios are consonant and
harmonious. Musical intervals represent the ratios of frequencies of two notes. For
example, an octave, which represents the ratio 2:1, is the interval between two notes,
one having double the frequency of the other. A perfect fifth is an interval represents
the ratio 3:2. There are many other intervals such as perfect fourth “4:3”, whole step
“9:8”, etc. In fact, our study here is about the first two intervals which are the most
consonant intervals.

Pythagorean scale used only octaves and perfect fifths. The problem we are

()
y
3
trying to solve comes from trying to find an integer solution to the equation 2 x = in
2
order to keep the scale finite. This equation has no integer solution except x = y = 0.
So, we need to approximate the solution using continued fractions.

() ()
y y
3 x 3
Now, 2x = implies that 2 y = .
2 2

x 3
So, =log 2( )= log 2(3)−1≈ 0.5849625007211562 and its continued fraction
y 2
expansion is [0, 1, 1, 2, 2, 3, 1, 5, 2, 23, 2, …]

1 3 7 24 31
The Kth convergents, 2 ≤ k ≤ 6 are: , , , ,
2 5 12 41 53

( ) 7
Taking the fourth convergent, we get 3 ≈ 2 12 =1.4983070768766815 .
2

7
The approximation implies that the octave consists of 12 semitones with a
12
perfect fifth equal to 7 semitones, “a semitone is the musical interval between two

40
adjacent notes in a 12-tone scale”. In fact, in western music, they use this
approximation, i.e., the octave is divided into 12 semitones. Other approximations are
24 31
inconvenient since and give 41 and 53 notes within an octave, respectively
41 63
1 3
which are too many notes. Moreover, and give 2 and 5 notes within an octave,
2 5
7
respectively which are too few notes. The percentage error results from choosing
12

as an approximation is: |( log (3 )−1 ) − 12 |

2
7
100 %=0.278508 % <0.3 % .
log 2 (3)−1

4.4 AN APPLICATION OF THE SOLAR SYSTEM

A cog wheel has a whole number of teeth round its rim that connect with (mesh
with) the teeth on another cog. The ratio of the two numbers of teeth governs the
speed ratio between the two cogs. Thus a cog with 10 teeth meshing with one with 50
means the 10-tooth cog will rotate 5 times quicker than the 50-toothed cog (but in the
opposite direction). So a cog can only revolve at a fixed fraction of the rate of any
other to which it is connected, and so on for a train of cogs (a gear train). Most
vehicles, from bicycles to cars and trains, use cogs like this in a gearbox to “change
gear”, that is, to keep the engine (or pedals) rotating in a narrow range of speeds but
allowing the wheels to go at increasingly faster or slower speeds.

41
 An application of continued fraction convergents is if we wish to make two cog
wheels where one rotates √ 2 times faster than the other, for example. Since √ 2
irrational, there is no cog mechanism that will give this exactly so we have to
approximate using fractions.

1 1 1
√ 2=1+ =1+ =1+
1+ √ 2 1+1+
1
2+
1
1+ √ 2 1+ √ 2

1 1
¿ 1+ =1+
1 1
2+ 2+
1 1
1+1+ 2+
1+ √ 2 1+ √ 2

1 1
¿ 1+ =1+
1 1
2+ 2+
1 1
2+ 2+
1 1
1+1+ 2+
1+ √2 1+ √2

a=√ 2=2.41421358 .. .

a 0=1

3 1
a 1= =1+
2 2

7 1
a 2= =1+
5 1
2+
2

17 1
a 3= =1+
12 1
2+
1
2+
2
41 1
a 4= =1+
29 1
2+
1
2+
1
2+
2

The convergents for √ 2 that we found above are:

42
 3 7 17 41 99
2 5 12 29 70

We could have 7 teeth on one cog and 5 on another, but it is difficult to get
efficient meshing with so few teeth so we would use perhaps 70 and 50. 17 and
12 teeth would give a closer approximation. If we allow ourselves up to 100
teeth on a cog, then the best approximation to √ 2 is given by 99 teeth and 70,
which gives an error of only 0.007%. Such fractions would be useful to know if
you were building a clockwork model of the solar system (an orrery) where we
wanted the planets to rotate accurately in relation to one another. For example
we would want the earth to spin around its own axis exactly 365.242199
times(“days”) in the time it takes it to turn exactly once around the sun (a
“year”), for example. Similarly, the moon must take 29.53059 “days” to rotate
exactly once around the earth. Such an accurate (mechanical) model is called an
orrery. Christian Huygens (1629-1695) used continued fractions when
constructing his orrery.

43
 CONCLUSION

In this dissertation, we discussed about continued fraction, properties of

continued fraction and periodic continued fraction. Also, an introduction to number
theory and some definition are discussed. Basic concepts which are used in our
dissertation are also discussed.

Also, we discussed about application of continued fraction in calendar

construction, Astronomy, Music and Solar system. So, this can be considered as a
first step towards my research.

44
 BIBLIOGRAPHY

[1] David M. Burton, Elementary number theory, MC Graw Hill education (India)

Edition 2012.

[2] K.C. Chowdhury, A first course in theory of numbers, Asian books Ltd, new Delhi

2004.

[3] Ivan Niven and H.S. Zuckerman, An introduction to the theory of Numbers, 3rd

edition, Wiley eastern Ltd, New Delhi, 1989.

[4] https://pi.math.cornell.edu/ gautam/continuedfraction.pdf

[5] https://r-knott.survey.ac.uk/Fibonacci/cfINTRO.htm|#section15

[6] https://pi.math.cornell.edu|m|sites|default/files/imported/Research/seniortheses/2007/

chengytthesis.pdf

45
46