Seminar

Integral function and Gamma Function

Deenbandhu Chhotu Ram University of

Science and Technology, Murthal

Submitted by : Medhavi
Class : M.Sc Mathematics 2nd year
Roll number : 20001553033

Submitted to : Dr. Nawneet Hooda

 INDEX
S. no. Topic Page
number
1 Introduction 3
2 Integral Function 4
3 Construction of 6
integral function

4 WEIERSTRASS 6
PRIMARY FACTOR

5 WEIERSTRASS 10
FACTORIZATION
THEOREM

6 M.L Theorem 14

7 Gamma function 18

8 SIMPLE PROPERTIES 22
OF GAMMA
FUNCTION

9 DUPLICATION 26
FORMULA OR
LEGENDER’S FORMULA

2
 INTRODUCTION

Many important properties of analytic functions are very difficult to

prove without use of complex integration. For instance, it is only
recently that it became possible to prove, without resorting to
complex integrals or equivalent tools, that the derivative of an
analytic function is continuous, or that the higher derivatives exist.
At present the integration-free proofs are, to say the least, much
more difficult than the classical proofs.
As in the real case we distinguish between definite and indefinite
integrals. An indefinite integral is a function whose derivative equals
a given analytic function in a region; in many elementary cases
indefinite integrals can be found by inversion of known derivation
formulas. The definite integrals are taken over differentiable or
piecewise differentiable arcs and are not limited to analytic
functions. They can be defined by a limit process which mimics the
definition of a real definite integral. Actually, we shall prefer to
define complex definite integrals in terms of real integrals. This will
save us from repeating existence proofs which are essentially the
same as in the real case. N aurally, the reader must be thoroughly
familiar with the theory of definite integrals of real continuous
functions. Very general theorems have their natural place in the
theory of analytic functions, but it must also be kept in mind that the
whole theory originated from a desire to be able to manipulate
explicit analytic expressions. Such expressions take the form of
infinite series, infinite products, and other limits. we deal partly with
the rules that govern such limits, partly with quite explicit
representations of elementary transcendental functions and other
specific functions.

3
INTERGRAL FUNCTION
A function which is regular in every finite region of z- plane is called
integral function.
In other words,
An integral function is an analytic function having no
singularity except at ∞ .

For eg : i) 𝑒 #$ is an integral function .

ii) the simplest function is polynomial.
As we know if a polynomial has n root then p(z) can be written as
product of linear factor as follows:
𝑧 𝑧 𝑧 𝑧
( ) ( )
𝑝 𝑧 = 𝑝 0 . ,1 − 0 ,1 − 0 ,1 − 0 … … … ,1 − 0
𝑧/ 𝑧# 𝑧1 𝑧4

This polynomial has zero at the point 𝑧/ , 𝑧# , … … … 𝑧4 .

• An integral function which is not polynomial may have an

$
infinite no. of zeros and the product ∏(1 − ) taken over
$7
these zeroes may be divergent.
∴ every integral function can’t be factorized as similar polynomial.

Remark:
1. An integral function may have no zero.
For eg : f(z)=𝑒 $ has no zero in ∅.
2. An integral function may have finite no. of zero
For eg : polynomial.
3. An integral function may have infinite no. of zeros
For eg : f(z) = sin 𝑧 .

4
THEOREM: The most general integral function with no zero f s
of the form 𝑒 =($) ; g(z) is itself an integral function.
Proof: - Let f(z) is a integral function with no zero then f’(z) is also
an integral function .

>?($)
so is .
>($)
$ >?(A)
Let F(z)= ∫$ 𝑑𝑡 ; where the integeral is taken along any path
B >(A)
from fixed point 𝑧E to the point z.
𝐹 (𝑧) = [log(𝑓(𝑡))]|$$B

𝐹 (𝑧) = log(𝑓(𝑧)) − log (𝑓(𝑧E ))

𝐹 (𝑧) + log(𝑓(𝑧)) = log (𝑓(𝑧E ))

𝐹 (𝑧) = 𝑒𝑥𝑝[𝐹 (𝑧) + log (𝑓(𝑧E ))]

= exp (𝑔(𝑧))

𝐹 (𝑧) = 𝑒 =($) ; where g(z)=F(z)+log(f(𝑧E )) which is an integral

function.

5
CONSTRUCTION OF INTEGRAL
FUNCTION:
WITH GIVEN ZEORS: -
If f(z) is an integral function with only a finite number of zeroes,
say( 𝑧/ , 𝑧# , … … … 𝑧4 ).
Then the function
>($)
($T$U )($T$V )………($T$7 )
is an integral function with no zeros.

Also, by theorem we know that most general form of an integral

function having no zero is 𝑒 =($) , where g(z) is integral function
again.
𝑓 (𝑧 )
𝑒 =($) =
(𝑧 − 𝑧/ )(𝑧 − 𝑧# ) … … … (𝑧 − 𝑧4 )

REMARK: -
To determine an integral function with an infinite no. of zeros , we
have an important theorem due to Weierstrass.

WEIERSTRASS PRIMARY FACTOR

The expression
𝐸(𝑧, 0) = (1-z)𝑒 E$
𝐸(𝑧, 1) = (1-z)𝑒/.$
$ VZ
𝐸(𝑧, 2) = (1-z)𝑒 $Y #
.
.
.
[V [\ [^
$Y V Y \ Y⋯Y ^
𝐸(𝑧 , 𝑝) = (1-z)𝑒 ; (p>0) are called Weierstrass
primary factor .

6
NOTE: - Each primary factor is an integral function with only
simple zero at z=1.

CASE 1) For |z|< 1

As we know
(𝑧 − 1)T/ = 1+z+𝑧 # + 𝑧 1 +………….

/
= 1 + 𝑧 + 𝑧 # + 𝑧 1 +……………..(*)
(/T$)

Integrate (*), we get

$V $\
𝐿𝑜𝑔 (1 − 𝑧) (−1) = 𝑧 + + +⋯
# 1
$V $\
𝐿𝑜𝑔 (1 − 𝑧) =−𝑧 − − −⋯
# 1
Now
[V [\
$Y V Y \ Y⋯
E (z, p) = (1-z )𝑒

[V [^
abc(/T$)Y($Y V Y⋯………….Y ^ )
=𝑒

$ ^eU $ ^eV $ ^e\

= 𝑒𝑥𝑝 d− − − − ⋯g
fY/ fY# fY1

$ ^eU $ ^eV $ ^e\

𝐿𝑜𝑔 (𝐸(𝑧, 𝑝)) = d− − − − ⋯g
fY/ fY# fY1

/
CASE 2) FOR |z|≤ #

From (1) we have

𝑧 fY/ 𝑧 fY# 𝑧 fY1
ilogj𝐸 (𝑧, 𝑝)ki = | − − − − ⋯|
𝑝+1 𝑝+2 𝑝+3

7
 |$|^eU |$|^eV
≤ + +⋯
fY/ fY#

fY/ / |$| |$|V

= |𝑧| d + + + ⋯g
fY/ fY/ fY#

≤ |𝑧|fY/ [1 + |𝑧| + |𝑧|# + |𝑧|1 ]

/ / /
= |𝑧|fY/ d! + + + g
# n o

/
ilogj𝐸 (𝑧, 𝑝)ki ≤ |𝑧|fY/ U
(/TV)

|$|^eU
= /Z = 2|𝑧|fY/
#

/
ilogj𝐸 (𝑧, 𝑝)ki ≤ |𝑧|fY/ where |z|≤
#
-------(2)

THEOREM →If < 𝑧/ , 𝑧# , 𝑧1 , … , 𝑧4 , … > be any sequence of

number whose only limit point is a point at infinity, then it is
possible to construct an integral function which is vanishes at each
point at {𝑧4 } and nowhere else.
Proof:
Let the given zeroes < 𝑧/ , 𝑧# , 𝑧1 , … , 𝑧4 , … > be arranged in
order of non-decreasing modulus.
𝑖. 𝑒 0 ≤ |𝑧/ | ≤ |𝑧# | ≤ |𝑧1 | ≤ ⋯ … … . ≤ |𝑧4 | ≤……….

Let |𝑧4 | = 𝑟4
Since, 𝑟4 is increasing indefinitely with n , we can find a sequence
of +ve integers 𝑝/ , 𝑝# , 𝑝1 , … … … … , 𝑝4 , ….
w f7
Such that the series ∑y
z{/ v x ---------(1)
w7
is cgt for all +ve value of r .
It is always possible to find such a sequence.

8
We can find {𝑝4 } = {n}

w /
Since ∑( 7) < ∑ is cgt series.
w #7

∴ By Weierstrass M-test

w
∑y
4{/( )
4
is cgt uniformly.
w7

$
Let 𝑓 (𝑧) = ∏y
4{/ 𝐸 ( , 𝑥4 − 1) ……….(2)
$7

$ /
This function has required property, for |𝑧4 |> 2|𝑧| , | | <
$7 #
.

$ $ f7 w f7
Then we have |log 𝐸 v , 𝑝4 − 1x| ≤ 2 | | =2 v x
$7 $7 w7

w
But ∑y
4{/( )
f7
is uniformly convergent.
w7

Hence, the series

$
∑y
4{/ log 𝐸 ( , 𝑝4 − 1) is uniformly convergent.
$7

For |𝑧4 | > 2 𝑅 𝑎𝑛𝑑 |𝑧| ≤ 𝑅 .

And also by Weierstrass test for the uniform convergence of an
infinite product to be
$
convergent. So, is the product ∏y
4{/ 𝐸( , 𝑝4 − 1) , |𝑧4 | ≥ 2 𝑅.
$7

Hence, p(z) is regular |𝑧| ≤ 2𝑅 and its only zeros in the region are
those of

9
 y
2
𝑓 (𝑧) = • 𝐸 , , 𝑝4 − 10 , |𝑧4 | > 2𝑅
𝑧4
4{/

𝑖. 𝑒 𝑡ℎ𝑒 𝑝𝑜𝑖𝑛𝑡 𝑧/ , 𝑧# , 𝑧1 , … … ….

Since, ℛ can be taken as large as please, therefore we conclude

that f(z) is regular function over ℂ. 𝑖. 𝑒 𝑓(𝑧) is integral function
which varnishes only at 𝑧/ , 𝑧# , 𝑧1 , … … ….
And nowhere else.
Hence proved

REMARK: The function f(z) is not uniquely defined.

WEIERSTRASS FACTORIZATION
THEOREM
Let f(z) be an entire function and entire function and let {𝑧4 } be
the non-zero, zeros of f(z) repeated according to their multiplicity,
suppose f(z) has a zero at z=0 of order in
(≥ 0) . Then there is an entire function g(z) and a sequence of
integer {𝑝4 } such that

y
𝑧
𝑓 (𝑧) = 𝑧 ‡ 𝑒 =($) • 𝐸( , 𝑝 − 1)
𝑧4 4
4{/
By the above theorem, {𝑝4 } can chose such that

10
 y
𝑧
ℎ(𝑧) = 𝑧 ‡ • 𝐸( , 𝑝 − 1)
𝑧4 4
4{/
has the same zeros of 𝑓(𝑧) with the same multiplicity it follows
>($)
that has removable singularity.
z($)
At 𝑧 = 0, 𝑧/ , 𝑧# , 𝑧1 , … … ….

>($)
Thus, is an entire function which has no zero. Since , C is
z($)
simply connected , then ∋ an entire function g(z) such that

>($)
= 𝑒 =($)
z($)

𝑓 (𝑧) = ℎ(𝑧) 𝑒 =($)

$
𝑓 (𝑧) = 𝑧 ‡ ∏y
4{/ 𝐸( , 𝑝4 − 1)𝑒 =($)
$‰

$
𝑓 (𝑧) = 𝑧 ‡ . 𝑒 =($) ∏y
4{/ 𝐸 ( , 𝑝4 − 1)
$‰

Hence proved

THEOREM: Every infinite function is meromorphic in the entire

z- plane is the quotient of the integral function.
Proof:
Let 𝜙(𝑧) be a function which is meromorphic in the entire z-
plane.
Then, we can find an entire function g(z) with the poles of 𝜙(𝑧) as
its zero.
Thus, the product 𝜙(𝑧). 𝑔(𝑧) is entire function say (𝑓(𝑧)).

11
 𝑓(𝑥)
∴ 𝜙(𝑧 ). 𝑔 (𝑧 ) = 𝑓 (𝑧 ) = = 𝜙(𝑧)
𝑔(𝑥)

>($)
∴ 𝜙(𝑧)(𝑚𝑜𝑟𝑜𝑟𝑝ℎ𝑖𝑐 𝑓𝑢𝑛𝑐𝑡𝑖𝑜𝑛) =
=($)
THEOREM: If f(z) is an integral function and 𝑓(0) ≠ 0 . Then
𝑓 (𝑧) = 𝑓 (0) 𝑝(𝑧) 𝑒 =($) , where p(z) is the product of primary
factor and g(z) is an entire function.
Proof:
We can form p(z) from the zero of 𝑓(𝑧)

•($)•
Let 𝜙(𝑧) =
f($)
This 𝜙(𝑧) is an integral function. Since, the poles of one term are
cancelled by the poles of other or zeros of other.

$ $ >?(A) f?(A)
Now, let 𝑔(𝑧) = ∫E 𝜙(𝑡)𝑑𝑡 = ∫E d − g 𝑑𝑡
>(A) f(A)

= log 𝑓 (𝑧) − log 𝑓(0) − log(𝑝(𝑧) +

log(𝑝(0)))

Since, p(z) is the product of primary factor.

⟹ 𝑝(0) = 1

⇒ logj𝑝(0)k = log 1 = 0

⇒ logj𝑝(0)k = 0
⇒ logj𝑓 (𝑧)k = 𝑔(𝑧) + logj𝑓(0)k + log(𝑝(𝑧))

𝑓(𝑧) = 𝑒 =(“)Yabcj>(E)kYabc(f($))

= 𝑒 c(”)Yabc( >(E) (f($))

= 𝑒 abc(>(E).f($)) . 𝑒 =($)

12
 𝑓(𝑧) = 𝑓(0)𝑝(𝑧)𝑒 =($)

Hence Proved

THEOREM: Let f(z) be an analytics function of z and suppose

that it has simple zero at the point 𝑧 = 𝑎/ , 𝑎# , 𝑎1 , … … … … then
f(z) can be expressed as:

[ ••([)
$ $Z
𝑓 (𝑧) = 𝑓 (0)𝑒 •([) ∏y
4{/(1 − )𝑒 –7 .
–7

Proof:
Since 𝑎/ , 𝑎# , 𝑎1 , … … … … are the simple zero of f(z).
∴ In the neighborhood of 𝑧 = 𝑎4 , 𝑓(𝑧) can be written as
𝑓 (𝑧) = (𝑧 − 𝑎4 )𝑔(𝑧) ; 𝑔(𝑎4 ) ≠ 0 ------------(1)
and analytic
Taking 𝑙𝑜𝑔 on both side side of --------------(1),
We get
log(𝑓(𝑧)) = log((𝑧 − 𝑎4 ). 𝑔(𝑧))

= log((𝑧 − 𝑎4 ) + log(𝑔(𝑧)) ------------(2)

Now differentiate it w.r.t z , we get

/ / =?($)
. 𝑓 ? (𝑧 ) = + ----------------(3)
>($) $T–7 =($)

𝑔′(𝑧)
Where, š𝑔(𝑧) is analytic at
𝑧 = 𝑎4 (∵ 𝑔(𝑎4 ) ≠ 0)

>?($)
Equation (3) shows that has simple pole at 𝑧 = 𝑎4 .
>($)

13
 𝑓? >?($)
Also, 𝑅𝑒𝑠 v š𝑓 , 𝑧 = 𝑎4 x = lim (𝑧 − 𝑎4 ) . .
$→–7 >($)

1 𝑔′(𝑧)
lim (𝑧 − 𝑎4 ) . ž + Ÿ
$→–7 (𝑧 − 𝑎4 ) 𝑔(𝑧)

𝑔′(𝑧)
lim 1 + (𝑧 − 𝑎4 ) .
$→–7 𝑔(𝑧)

>•
𝑅𝑒𝑠 v , 𝑧 = 𝑎4 x = 1 + 0 = 1
>

RECALL : M.L Theorem

STATEMENT: - Let f(z) be a meromorphic function whose only
singularity except at ∞ are the poles 𝑎/ , 𝑎# , 𝑎1 , … … … … 𝑎4
where they are arrange in ascending order

0 ≤ |𝑎/ | ≤ |𝑎# | ≤ |𝑎1 | ≤ |𝑎n | ≤ ⋯ … … … … ≤ |𝑎4 |

And let Let 𝑏/ , 𝑏# , 𝑏1 , … … … … . 𝑏4 respectively be the residue at

the poles 𝑎/ , 𝑎# , 𝑎1 , … … … … 𝑎4 .

Let {𝑐4 } be a sequence of closed contour such that 𝑐4 includes

𝑎/ , 𝑎# , 𝑎1 , … … … … 𝑎4 and no other poles
Also suppose that
(i) The minimum distance 𝑅4 of points of 𝑐4 from the origin
tends
to ∞ 𝑎𝑠. 𝑛 → ∞
(ii) F(z) is bounded on 𝑐4 𝑖. 𝑒 |𝑓(𝑧)| ≤ 𝑀 𝑜𝑛 𝐶4 . The f(z) can be
expanded as:
y
1 1
𝑓(𝑧) = 𝑓 (0) + £ 𝑏4 ¤ + ¥
𝑧 − 𝑎4 𝑎4
4{/

14
 >?($)
Now satisfies the M.L theorem.
>($)
>?
Therefore, can be written as
>

>?($) >?(E) / /
= + ∑y
4{/ 1. d + g------------(4)
>($) >(E) $T–7 –7

>?($)
Now integrate equation (3) w.r.t z. from z=0 to z=2
>($)
along a path not passing through any pole.
y
𝑓′(0) 1
log 𝑓(𝑧) |E$ = . 𝑧 + £ log(𝑧 − 𝑎4 ) + . 𝑧 |E$
𝑓(0) 𝑎4
4{/

𝑓′(0)
log 𝑓(𝑧) − log 𝑓(0) = .𝑧
𝑓(0)
y
1
+ £ ¤log(𝑧 − 𝑎4 ) − log(−𝑎𝑛 ) + 𝑧¥
𝑎4
4{/

y
𝑓(𝑧) 𝑓′(0) 𝑧 − 𝑎4 1
log = . 𝑧 + £ ¤log , 0 + 𝑧¥
𝑓(0) 𝑓(0) −𝑎4 𝑎4
4{/

y
𝑓(𝑧) 𝑓′(0) 𝑧 $
log = . 𝑧 + 𝑙𝑜𝑔 •(1 − )𝑒 Z–7
𝑓(0) 𝑓(0) 𝑎4
4{/

• y
𝑓(𝑧) > (E).$ 𝑧 $
log = log 𝑒 >(E) + log (• ,1 − 0 . 𝑒 Z–7 )
𝑓(0) 𝑎4
4{/

• y
𝑓(𝑧) > (E).$ 𝑧 $
log = log 𝑒 >(E) . • ,1 − 0 . 𝑒 Z–7 )
𝑓(0) 𝑎4
4{/

15
 y
>• (E).$ 𝑧 $
⇒ 𝑓 (𝑧 ) = 𝑓 (0). 𝑒 >(E) . • ,1 − 0 . 𝑒 Z–7
𝑎4
4{/

Hence proved

$V
EXAMPLE: show that.
¦§¨ $
$
= ∏y
4{/(1 − 4V © V
)

Solution:
¦§¨ $
Let 𝑓 (𝑧) =
$

>?($) $ ª«¬ $T¦§¨ $ $ $ -b¦ $T¦§¨ $

Now, = . =
>($) $V ¦§¨ $ $ ¦§¨ $

>?($)
Now, the pole of are given by z sin z =0
>($)
⇒ 𝑧 = 0 𝑜𝑟 sin 𝑧 = 0
⇒ 𝑧 = 𝑛𝜋 ; 𝑛 ∈ Ζ

But z=0 is a removable singularity.

>?($)
So, has poles at z=n𝜋 ; 𝑛 = ±1 , ±2 , ±3 , ±4 … … ….
>($)
Or 𝑛 ∈ 𝑧 − {0}.

𝑓′(𝑧)
Now, š𝑓(𝑧) satisfies the condition of M.L theorem.

y
$
>?(E)
š>(E) 𝑧 $Z–
⇒ 𝑓(𝑧) = 𝑓 (0). 𝑒 •(1 − )𝑒 7
𝑎4
4{/

16
 sin 𝑧 sin 𝑧
𝑓 (𝑧 ) = ⇒ 𝑓 (0) = lim =1
𝑧 $→E 𝑧

𝑧 cos 𝑧 − sin 𝑧
?(
)
𝑓 𝑧 =
𝑧#
?( )
cos 𝑧 − sin 𝑧 0
⇒ 𝑓 0 = lim 𝑧. = , 0 𝑓𝑜𝑟𝑚
$→E 𝑧# 0

By L’ Hospital Rule, we get

sin 𝑧 + cos 𝑧 − cos 𝑧 sin 𝑧
𝑓 ? (0) = −𝑧 = − =0
2𝑧 2
⇒ 𝑓 ? (0) = 0

𝑓′(0) 0
⇒ = =0
𝑓(0) 1
y
𝑧 $
⇒ 𝑓 (𝑧) = 1. 𝑒 E.$ • ,1 − 0 . 𝑒 Z–7
𝑎4
4{/

$ $Z
= ∏y
4{/ v1 − x.𝑒 –7
–7

y y
𝑠𝑖𝑛𝑧 𝑧 $ 𝑧 T$
= • v1 − x . 𝑒 ⁄4© . •(1 + )𝑒 ⁄4
𝑧 𝑛𝜋 𝑛𝜋
4{/ 4{/

y
𝑧# $ $
= • µ1 − # # ¶ . 𝑒 4©T4©
𝑛 𝜋
4{/

¦§¨ $ $V
= ∏y
4{/ v1 − x
$ 4V © V

17
 GAMMA FUNCTION
Here, we shall construct a function called Gamma function or
Euler Gamma function which is meromorphic with poles at non-
positive integer 𝑖. 𝑒 𝑧 = 0 , −1 , −2 , −3 , … … …

There are two natural approach to construct Gamma function.

1) By Weierstrass product
2) 2) Mellin integral

Certain properties are clear from one definition but not from
other, although the two definition give the same function.
We start with the former approach which involves more
algebraic properties of Gamma function.
For this, we introduce function which have only -ve zeroes.

STEP 1): The simplest function of the type is:

y
𝑧 T$
𝐺 (𝑧) = • v1 + x . 𝑒 ⁄4 − − − −(1)
𝑛
4{/
Then
y
𝑧 $
𝐺(−𝑧) = • v1 − x . 𝑒 ⁄4
𝑛
4{/

y
𝑧# 𝜋𝑧
( ) ( )
∴ 𝑧. 𝐺 𝑧 . 𝐺 𝑧 = 𝑧 • µ1 − # ¶ = sin − − − − − (2)
𝑛 𝜋
4{/
Now

18
 y
𝑧 − 1 T($T/)Z
𝐺 (𝑧 − 1) = • ,1 + 0𝑒 4 − − − − − −(3)
𝑛
4{/

The zeroes of G(z-1) given by the factor

𝑧−1
,1 + 0=0
𝑛

𝑧−1
⇒ = −1
𝑛

⇒ 𝑧 = −𝑛 + 1 ; 𝑓𝑜𝑟 𝑛 = 1, 2 , 3, 4 , … … ….
⇒ 𝑧 = 0, −1 , −2 , −3 , −4 , … … …
Thus, G(z-1) has same zero as G(z) and one addition zero at z=0
(simple zero at z=0)
So, By Weierstrass factorization theorem, we get
𝐺 (𝑧 − 1) = 𝑧 𝑒 =($) . 𝐺 (𝑧) − − − − − (4)
Where g(z) is an integral function.

STEP 2) Use (1) and (3) into (4) we get

y y
𝑧 − 1 T($T/)Z 𝑧 T$
• ,1 + 0𝑒 4 = 𝑧 𝑒 =($) • v1 + x . 𝑒 ⁄4 − −(5)
𝑛 𝑛
4{/ 4{/

Now take 𝑙𝑜𝑔 on both side (to determine g(z))

y y
𝑧 − 1 T($T/)Z 𝑧 T$
⇒ 𝑙𝑜𝑔 • ,1 + 0𝑒 4 = 𝑙𝑜𝑔 𝑧 + 𝑔 (𝑧) + £ log dv1 + x . 𝑒 ⁄4 g
𝑛 𝑛
4{/ 4{/

y y
𝑧−1+𝑛 −(𝑧 − 1) 𝑧 −𝑧
⇒ £ ¤log , 0+( )¥ = 𝑙𝑜𝑔𝑧 + 𝑔(𝑧) + £ log (1 + )( )
𝑛 𝑛 𝑛 𝑛
4{/ 4{/

19
 ∞ ∞
𝑧−1+𝑛 −(𝑧 − 1) 𝑧 −𝑧
⇒ £ dlog v x+( )g = 𝑙𝑜𝑔𝑧 + 𝑔(𝑧) + £ dlog (1 + )( )g
𝑛 𝑛 𝑛 𝑛
𝑛=1 𝑛=1

Diff both side 𝑤. 𝑟. 𝑡 𝑧, we get

y y
𝑛 1 1 1 1 1
⇒ £¤ . − ¥ = + 𝑔 ? (𝑧 ) + £ ¤ − ¥ − −(6)
𝑧−1+𝑛 𝑛 𝑛 𝑧 𝑧+𝑛 𝑛
4{/ 4{/

y y
1 1 1 1 1
⇒ £¤ − ¥ = + 𝑔 ? (𝑧 ) + £ ¤ − ¥
𝑧−1+𝑛 𝑛 𝑧 𝑧+𝑛 𝑛
4{/ 4{/

SETP 3) Replacing n by (n+1); the series (L.H.S of equation (6))

can be written as .
y y
1 1 1 1
£¤ − ¥= £ ¤ − ¥
𝑧−1+𝑛 𝑛 𝑧+𝑛 𝑛
4{/ 4Y/{/

y y
1 1 1 1
£¤ − ¥ =£¤ − ¥
𝑧−1+𝑛 𝑛 𝑧+𝑛 𝑛
4{/ 4{E

y
1 1 1
= , − 10 + £ ¤ − ¥
𝑧 𝑧+𝑛 𝑛
4{/

y y
1 1 1 1 1
= , − 10 + £ ¤ − ¥+ £¤ − ¥
𝑧 𝑧+𝑛 𝑛 𝑛 𝑛+1
4{/ 4{/

y
1 1 1
= , − 10 + £ ¤ − ¥+ 1
𝑧 𝑧+𝑛 𝑛
4{/

20
 y
1 1 1
= , 0+ £¤ − ¥ − − − − − (7)
𝑧 𝑧+𝑛 𝑛
4{/

By using (7) into (6) we get

y y
1 1 1 1 1 1
, 0+ £¤ − ¥ = + 𝑔 ? (𝑧 ) + £ ¤ − ¥
𝑧 𝑧+𝑛 𝑛 𝑧 𝑧+𝑛 𝑛
4{/ 4{/
⇒ 𝑔 ? (𝑧 ) = 0 ∀ Ζ
⇒ 𝑔(𝑧) 𝑖𝑠 𝑎 𝑐𝑜𝑛𝑠𝑡𝑎𝑛𝑡 𝑓𝑢𝑛𝑐𝑡𝑖𝑜𝑛 .

𝑙𝑒𝑡 𝑔(𝑧) = 𝑟 (𝑠𝑎𝑦 )

STEP 4): Hence, equation (4) can be written as follows
𝐺 (𝑧 − 1) = 𝑧𝑒 w 𝐺(𝑧) − − − −(8)

To determine r, we put (z=1) into (8) , we get

𝐺 (0) = 1𝑒 w 𝐺 (1)
From (1)
𝐺 (0) = 1
¿U
/
And 𝐺 (1) = ∏y
4{/(1 + )𝑒 7
4

¿U
Tw /
𝑒 = ∏y
4{/ v1 + x 𝑒 7 − − − −(9)
4
1 1 1
−𝑟 = lim ¤log(1 + 𝑛) − (1 + + + − − + )¥
4→y 2 3 𝑛

1 1 1
𝑟 = lim ¤,1 + + + − − − − + 0 − log 𝑛¥
4→y 2 3 𝑛

This r is called Euler’s constant.

Its approximate value is = 57721.6

21
 GAMMA FUNCTION
The ⌈ ( 𝐺𝑎𝑚𝑚𝑎 )function is defined as
1
Â𝑧 = Ã$
𝑧𝑒 𝐺(𝑧)
y
𝑒 Tw$ 𝑧 $
Ä𝑧 = •(1 + )T/ 𝑒 ⁄4 − − − −(10)
𝑧 𝑛
4{/

Note that ⌈𝑧 is well defined in the whole complex plane except z=

0 ,-1 ,-2 ,-3 ,……… which are the simple pole of the function
Gamma.
Hence, Gamma function is a meromorphic function whose only
pole at non-positive integers.

SIMPLE PROPERTIES OF
GAMMA FUNCTION
The following are the simple property of Gamma function: -

(a) ⌈𝑧 + 1 = 𝑧⌈𝑧 𝑜𝑟 ⌈𝑧 = (𝑧 − 1)⌈𝑧 − 1

22
 𝑒 T($T/)Æ 𝑒 T$Æ . 𝑒 Æ
Å𝑧 − 1 = =
(𝑧 − 1)𝐺(𝑧 − 1) (𝑧 − 1). 𝑧 𝑒 Æ 𝐺(𝑧)

1
(𝑧 − 1) Â𝑧 − 1 = = ⌈𝑧
𝑧 𝑒 $Æ 𝐺(𝑧)

⇒ ⌈𝑧 = (𝑧 − 1)⌈(𝑧 − 1)

Now replace z by (z+1) ⇒ ⌈𝑧 + 1 = 𝑧⌈𝑧

Hence Proved

©
(b) Ç1 − 𝑧 . Ç𝑧 =
¦§¨ ©$

As we know ⌈𝑧 + 1 = 𝑧⌈𝑧
⇒ ⌈𝑧 + 1 = −𝑧⌈−𝑧

⇒ ⌈𝑧 + 1 . ⌈𝑧 = −𝑧⌈−𝑧 . ⌈𝑧

/ /
= −𝑧. .
T$ Ã ¿[È É(T$) $Ã ÊÈ É ($)
/
=
$ Ã ¿[Ëe[Ë É($)É(T$)
/
=
$ É($) É(T$)
But
$ T$⁄
𝐺 (𝑧) = ∏y
4{/(1 + )𝑒
4
4

$ $⁄
⇒ 𝐺 (−𝑧) = ∏y
4{/(1 − ) 𝑒
4
4

$V
⇒ 𝐺 (𝑧) 𝐺(−𝑧) = ∏y
4{/(1 − 4V )
And we know that

23
 y
sin 𝑧 𝑧#
= •(1 − # # )
𝑧 𝜋 𝑛
4{/

y
sin π 𝑧 𝑧#
= •(1 − # )
𝜋𝑧 𝑛
4{/

sin 𝜋𝑧
⇒ = 𝐺(𝑧)𝐺(−𝑧)
𝜋𝑧

sin 𝜋𝑧
⇒ = 𝑧 . 𝐺(𝑧)𝐺(−𝑧)
𝜋
𝜋
⇒ Ç(1 − 𝑧) Ç𝑧 =
sin 𝜋𝑧

Hence Proved

(c) Í1Z2 = √𝜋

©
As we know: Ç(1 − 𝑧) Ç𝑧 = − − − − − −(∗)
¦§¨ ©$

Now replace z and 1Z2 into (*) , yield

𝜋
Í1Z2 . Ð1Z2 = 𝜋
sin
2
#
jÍ1Z2k = 𝜋

Í1Z2 = √𝜋

Hence proved

24
 (II) ⌈(𝑛 + 1) = 𝑛!

Since ⌈1 = 1
Further if n is a +ve integer, then using equation
⌈𝑧 + 1 = 𝑧⌈𝑧 , we get
⌈2 = ⌈1 = 1 𝑎𝑛𝑑 ⌈3 = 2⌈2
= 2(1)
= 2!
⌈4 = 3⌈3
= 3.2.1
=3! ………………………….. so on

Finally, we get ⌈𝑛 + 1 = 𝑛!

Or another approach: -
⌈(𝑛 + 1) = 𝑛!
For n= 0
⌈1 = 0! = 1
⇒ ⌈1 = 1
Let us suppose it is true for n= n
⇒ ⌈𝑛 + 1 = 𝑛!
Now we have to show for n= n+1
(to show). ⇒ ⌈𝑛 + 2 = (𝑛 + 1)!
⇒ ⌈𝑛 + 2 = (𝑛 + 1)⌈𝑛 + 1
= ⌈(𝑛 + 1) . 𝑛!
= (𝑛 + 1)!
Hence, by principle of Mathematical induction it is true for all
+ve integer n.

25
DUPLICATION FORMULA OR
LEGENDRE’S FORMULA
/ é
§ Show that Ç𝑧 . Ç(𝑧 + ) = ⌈2𝑧 .
# #V[¿U

Proof:
y
𝑒 TÆ$ 𝑧 $
Â𝑧 = •(1 + )T/ 𝑒 ⁄4 − − − −(1)
𝑧 𝑛
4{/
Taking log both sides, we get
y
𝑒 TÆ$ 𝑧 T/ $⁄
(⌈ )
log 𝑧 = 𝑙𝑜𝑔 Ñ •(1 + ) 𝑒 4 Ò
𝑧 𝑛
4{/

y
𝑧 T/ $⁄
= −𝛾𝑧 + £ log ¤v1 + x 𝑒 4 ¥ − log 𝑧
𝑛
4{/
y
𝑧 𝑧
𝑙𝑜𝑔⌈𝑧 = −𝛾𝑧 + £ d− log v1 + x + g − 𝑙𝑜𝑔𝑧
𝑛 𝑛
4{/
Differentiate both side with respect to ‘z’ , yield
y
⌈𝑧? −𝑛 1 1 1
= −𝛾 + £ , . + 0−
⌈𝑧 𝑧+𝑛 𝑛 𝑛 𝑧
4{/

y
−1 1 1
= −𝛾 + £ ¤ + ¥−
𝑧+𝑛 𝑛 𝑧
4{/
Again differentiate w.r.t z, yield
y
𝑑 ⌈′(2) 1 1
µ ¶ = 0 + £( + 0) +
𝑑𝑧 ⌈(2) (𝑧 + 𝑛 )# 𝑧#
4{/

26
 y
𝑑 ⌈′(2) 1 1
⟹ µ ¶=£ +
𝑑𝑧 ⌈(2) (𝑧 + 𝑛)# 𝑧 #
4{/

y
𝑑 ⌈′(2) 1
⟹ µ ¶=£
𝑑𝑧 ⌈(2) (𝑧 + 𝑛)#
4{E

Similarly, By similar process for replacing z by 2z, yield

y
𝑑 ⌈′(2) 1
µ ¶=2£
𝑑𝑧 ⌈(2) (2𝑧 + 𝑛)#
4{E

Now consider the term:

1 y y
𝑑 ⌈′(2) 𝑑 Ç′(𝑧 + 2) 1 1
µ ¶+ ( )=£ +£
𝑑𝑧 ⌈(2) 𝑑𝑧 Ç(𝑧 + 1) (𝑧 + 𝑛)# 1
(𝑧 + + 𝑛)#
2 4{E 4{E 2
/ /
= 4 ∑y
4{E + ∑y
4{E
(#$Y4#)V (#$Y/Y#4)V
y
1
=4£
(2𝑧 + 𝑚)#
‡{E
Here,

2𝑛 𝑖𝑓 𝑚 𝑖𝑠 𝑒𝑣𝑒𝑛
𝑚=Ô
(2𝑛 + 1) 𝑖𝑓 𝑚 𝑖𝑠 𝑜𝑑𝑑

Ö ⌈?(#$)
= 2. d ⌈(#$) g
Ö$
Therefore:
1
𝑑 ⌈′(2) 𝑑 Ç′(𝑧 + ) 𝑑 ⌈′(2𝑧)
µ ¶+ ( 2 ) = 2. ž Ÿ
𝑑𝑧 ⌈(2) 𝑑𝑧 Ç(𝑧 + 1) 𝑑𝑧 ⌈(2𝑧)
2
Integrate 2 time, we get

27
 1 log (⌈2𝑧)
𝑙𝑜𝑔 Â𝑧 + log (Â(2 + )) = 2 = log⌈2𝑧 + 𝑎𝑧 + 𝑏
2 2

1
log ,Â𝑧 . Â,𝑧 + 00 = 𝑙𝑜𝑔⌈2𝑧 + 𝑙𝑜𝑔𝑒 (–$Y×)
2

= log( ⌈2𝑧. 𝑒 (–$Y×) )

1
⇒ Â𝑧 Â(𝑧 + ) = Í2𝑧 . 𝑒 –$Y× − − − −(𝐴)
2

But z=1Z2 in (A)

1 –
⇒ Â . Í1 = Í1 𝑒 Z#Y×
2

1 –
⇒ Â = 𝑒 Z#Y×
2
–
⇒ √𝜋 = 𝑒 Z#Y×

And again, put z=1 into (A), we get

1
⇒ Â1 + . Í1 = Í2 . 𝑒 –Y×
2

3
⇒ Â . Í1 = Í2 . 𝑒 –Y×
2

1 1
⇒ Â = 𝑒 –Y×
2 2

√𝜋
⇒ = 𝑒 –Y×
2
𝑎 1
⇒ + 𝑏 = logj√𝜋k = log(𝜋)
2 2

28
 ⌈3
𝑎 + 𝑏 = log( )
2

1 1
= log ( Â )
2 2

1
= 𝑙𝑜𝑔 + 𝑙𝑜𝑔√𝜋
2

1
𝑎+𝑏 = 𝑙𝑜𝑔𝜋 − 𝑙𝑜𝑔2
2
𝑎
⇒ = −𝑙𝑜𝑔2
2

⇒ 𝑎 = −2𝑙𝑜𝑔2
1
⇒ 𝑏 = 𝑙𝑜𝑔𝜋 + 𝑙𝑜𝑔2
2
1
⇒ Â𝑧 . Â𝑧 + = Í2𝑧 . 𝑒 –$Y×
2
/ abc ©
T#$ Ù«=#Yabc #Y
= Â2𝑧 . 𝑒 #

/ abc ©
Ù«=#¿V[ Yabc #Y
= Â2𝑧 . 𝑒 #

/ abc ©
Ù«=(#¿V[eU )Y
= Â2𝑧 . 𝑒 #
¿V[eU )
= Í2𝑧 . 𝑒 Ù«=(# + 𝑙𝑜𝑔√𝜋

1
Â𝑧 . Â𝑧 + = ⌈2𝑧 . 𝜋. 2T#$Y/
2

1 √𝜋
⇒ Â𝑧 . Â𝑧 + = #$T/ . ⌈2𝑧
2 2

29
This is required duplication formula.

References:
1. H.A Priestly, Introduction to Complex Analysis,
Clarendon Press Oxford.
2. L.V Ahlfors, Complex analysis, McGraw-Hill.
3. J.W Brown and R.V Churchill, Complex variable and
applications, McGraw Hill.

30