ON SOME APPLICATIONS OF THE VOLONTERIOS TRANSFORM: SERIES

DEVELOPMENT OF TYPE Nk+M AND MATHEMATICAL CONNECTIONS WITH

SOME SECTORS OF THE STRING THEORY

Odoardo Volonterio1, Michele Nardelli2

Politecnico di Milano
Piazza Leonardo da Vinci, 32
20133 Milano, Italy
2

Dipartimento di Scienze della Terra

Universit degli Studi di Napoli Federico II, Largo S. Marcellino, 10
80138 Napoli, Italy

Abstract
In this work we have described a new mathematical application concerning the discrete and the analytic functions:
the Volonterios Transform (V Transform) and the Volonterios Polynomial. We have descrive various mathematical
applications and properties of them, precisely the series development of the type Nk+M. Furthermore, we have
showed also various examples and the possible mathematical connections with some sectors of Number Theory and
String Theory.

Index

ON SOME APPLICATIONS OF THE VOLONTERIOS TRANSFORM: SERIES DEVELOPMENT OF TYPE Nk+M AND
MATHEMATICAL CONNECTIONS WITH SOME SECTORS OF THE STRING THEORY ....................................................... 1
VOLONTERIOS TRANFORM GENERALIZED AND SERIES DEVELOPMENT OF TYPE Nk+M ............................................... 4
Generalized definition of transform V .................................................................................................................... 6
Generalized definition of inverse transform V. 6
RELATION BETWEEN THE VOLONTERIOS TRANFORM, OF LAPLACE AND ZETA.8
Definition and properties of the Volonterios polynomials10

Transformed V of a discrete periodic function 23

SERIES DEVELOPMENT OF THE TYPE Nk+M ......................................................................................................... 26
Assumption . ......................................................................................................................................................... 26
Proof..................................................................................................................................................................... 27
CASE 1 (N < M): ................................................................................................................................................. 31
CASE 2 (N = M) .................................................................................................................................................. 33
CASE 3 (N > M) .................................................................................................................................................. 34
PARTICULAR CASES............................................................................................................................................... 35
Case with

........................................................................................................................................ 35

Case with

........................................................................................................................ 37

EXAMPLES ............................................................................................................................................................ 37
Example 1 ......................................................................................................................................................... 37
Example 2 ......................................................................................................................................................... 38
Example 3 ......................................................................................................................................................... 39
Example 4 ......................................................................................................................................................... 40
ON SOME MATHEMATICAL CONNECTION WITH SOME SECTORS OF STRING THEORY..42
References..56

VOLONTERIOS TRANFORM GENERALIZED AND SERIES DEVELOPMENT OF TYPE

Nk+M
Definition 1 (transformed generalized V)
The transform V of a discrete function
is an analytic function of a real variable (or complex) through which it is
possible to pass from the world of discrete or finite mathematics in the world of differential mathematics.
The transformed V provides an overview higher than it can provide a generating function.
The transformed canonical is distinguished from generalized because its existence is based on continuous functions
and infinitely differentiable for t = 0 while the generalized is based on a continuous function and infinitely
differentiable at
(where for
we obtain obviously the transform canonical).
The properties of transformation and anti-transformation of the transform V are independent from the fact that we
consider the transformed canonical or generalized.

Definition 2 (inverse generalized transform V)

The inverse transform V of an analytic function
of a real variable t continues in the zero and infinite times
differentiable at t = 0 (canonical) or at t = x (generalized) (in other words a function
developable in MacLaurin
or Taylor series ) is a discrete function
defined in
through which it is possible the transition from the
differential world to the world of discrete or finite mathematics.

DEFINITION OF TRANSFORM V
Let

a discrete function, then we can define the transformation

as follows:

(1.1)

CONDITION OF EXISTENCE AND UNIQUENESS OF THE TRANSFORM V

To ensure the condition of existence of the transform must be ensured the following relation:
(a)

where

is the radius of the convergence while e is the Eulero-Nepero constant.

The relation (a) is a necessary condition that has been demonstrated exploiting the condition of the root of CauchyHadamard while the condition of uniqueness can be attributed to the properties of series of powers where e is the
Euler-Nepero constant.

Definition 1 of inverse transform V

We define with inverse transform of

the discrete function

obtained by the following definition:

(b)

Definition 2 of inverse transform V

Or by the following formula alternative to the (b):

(c)

where necessary and sufficient condition because (c) is valid is that is satisfied the condition
radius of convergence (a).

where

is the

Definition 3 of inverse transform V

(1.2)

or, for k N k 0 (see c):

T 1 V t , k :

k 1 R V ei cosk d

(d)

Generalized definition of transform V

As follows, we define the generalized transformed

1.3

Generalized definition of inverse transform V

We define as generalized inverse transform of

the discrete function

obtained by the following definition:

1.4

This definition is particularly useful in all those cases where the function
Another alternative definition is the following:

can not exist in 0.

1.5
6

Fundamental properties of the transform V

Among the generalized transform and the canonical transform, is useful to keep in mind the following identity:
1.6

1.7

1.8

RELATION BETWEEN THE VOLONTERIOS TRANFORM, OF LAPLACE AND ZETA

We consider the following definitions of Gamma function, Laplaces Transform and Zeta Transform:

1.9

we have:
1.10

thence:

1.11

1.12

but, for the left-hand side we note that:

1.13
thence:
1.14

while, for the right-hand side putting:

1.15

putting

we have:
1.16

1.17

vice versa
1.18

more generally, in the case of generalized transform V, we have:

1.19

1.20

ADDITIONAL DEFINITIONS
In order to read and interpret tables in complete sense clarifications is needed on the functions and abbreviations
that have been introduced and also will be essential of the examples that follow after the tables. In any case, before
proceeding to the list of transformations is useful to consider the following relations, definitions and functions.

Definition of operator
With the symbol

we define the following operator:

(1.21)

where its application iterated n times on a determined function

in the following way

is expressed by the operator of the application

for example, we consider V t sin t ,thence:

Definition and properties of the Volonterios polynomials

(1.22)

other formulas to determine the polynomial

(see TF N 4 and 20) are the following:

(1.23)

or by the following recursive formulas:

(1.24)

(1.25)

or:
(1.26)

From the above definition, by the Volonterios transform:

10

(1.27)

Definition of Bernoullis Polynomial:

The Bernoullis polynomial (see TF N 21) is:
(1.28)

The generating function of the Bernoullis polynomial is:

(1.29)

Definition of Euleros Polynomial:

Euleros Polynomial (see TF N 22):
(1.30)

The generating function of the Euleros polynomial is:

(1.31)

Definition of Laguerres Polynomial:

Laguerres Polynomial (see TF N 15 and N 23)
(1.32)

Definition of the Bessels Polynomial of the first kind:

(1.33)

Definition of the Hermites Polynomial

(1.34)

11

PROPERTIES OF THE TRANSFORM V

PF
N.

Function

Definition

1
2
3
4

6
7
8
9
10
11
12
13
14
15

16
17
18
19

12

TRANSFORMATION V OF SOME KNOWN FUNCTIONS

TF
N.

Function

Definition

1
2

(idem for 1)

3
4
5
6
7
8
9
10
11
12
13
14
15
16
17

18

19
20
21
22

13

TRANSFORMATION V OF SOME KNOWN FUNCTIONS

TF
N.

Function

Definition

23
24
25
26
27

28

with

29

14

FUNDAMENTAL PROPERTIES OF THE INVERSE TRANSFORM V

PA
N.

Function

Definition

1
2
3
4
5

10
11
12
13
14

15

INVERSE TRANSFORM OF KNOWN FUNCTIONS

AF
N.

Function

Definition

2
3
4
5
6
7
8
9
10
11
12
13
14
15

GENERALIZED TRANSFORMATION OF NOTE FUNCTIONS

GTF
N.

Function

Definition

1
2
3
4
5
6
7
8
16

GENERALIZED TRANSFORMATION OF NOTE FUNCTIONS

GTF
N.

Function

Definition

9
10
11
12
13
14
15
16
17

18

19
20
21
22
23
24
25
26
27

28

with

17

GENERALIZED TRANSFORMATION OF NOTE FUNCTIONS

GTF
N.

Function

Definition

29
30

18

GENERALIZED INVERSE TRANSFORMATION OF NOTE FUNCTIONS

GAF
N.

Function

Definition

3
4
5
6

8
9
10
11
12
13
14

15

16

19

From the GAF n.16, we observe:

1.35
Where, especially, if we want to use this property we must remember to run the rest of the calculations with a seed
equivalent based on the properties (1.6), (1.7) and (1.8).
1.36

1.37

20

Example
PROBLEM
Solve the following equation to the finite difference of the 2nd order.

(1.38)

Now, to solve such a simple equation to the finite difference of the second order homogeneous with constant
coefficients may be used various methods, including the method of the generating function and the method using
the transform realized here.
*
*
*
*
*
SOLUTION
a) METHOD OF THE GENERATING FUNCTION
We consider the following generating function:

(1.39)

(1.40)

(1.41)

(1.42)

(1.43)
taking into account the following observations:
(1.44)

From which we get the new generating function (have already been considered the initial conditions):
(1.45)

21

b)RESOLVING METHOD BY TRANSFORMED V

Calling with T this transformation from the variable
, we obtain:

to the variable

and placing

with

(1.46)

(1.47)

The characteristic equation associated is:

(1.48)

i.e., the solution of the differential equation with the initial settings is:
(1.49)

where now anti-transform, obtaining the solution searched:

(1.50)

22

Transformed V of a discrete periodic function:

Let

be a particular discrete function in which is valid the following relation:

1.51

Thanks to the Fourier series in the discrete domain:

1.52

we can apply the transformation V to the (1.52) as in (1.53):

1.53

for the transformations tables, we obtain:

1.54

1.55

1.56

Now replacing

of the (1.52) in the eq. (1.56) we obtain the eq. (1.57):

1.57

1.58

23

thence expanding the term

,we obtain:

1.59

Now, for the same definition of transformed

we have:

1.60

i.e. for

thence

namely:

1.61

1.62

from which we deduce the following relations:

1.63

1.64

In the particular case where the period N of the periodic discrete function is very large or even tending to infinity, we
proceed as follows:

24

where the step is

namely

thence rewrite the (1.63) as follows:

1.65

and for

we have

thence rewrite the eq. (1.65) as follows:

(1.66)

i.e.:

(1.67)

Given that

is the transformed

of

, we obtain the equality:

(1.68)

from which, we have that:

(1.69)

(1.70)

25

SERIES DEVELOPMENT OF THE TYPE Nk+M

Assumption
The intent is to find a form equivalent to the infinite sum below with

and

(2.1)

We will solve this problem by using the Volonterios Transform.

In these proofs and examples we will consider the convergence problem inherent in the solutions and criteria
adopted.

26

Proof
Consider the following examples with
and
needed to understand the proof that will follow, where
is a discrete periodic function of value 1, of period shifted by M.
Table 1

10

10

13

16

19

22

25

28

31

34

Table 2
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35
0 0 0 0 1 0 0 1 0 0

Thence for the (2.0) and to the Tables 1 and 2 where the relation between
following Table:

and

is

we obtain the

Table 3

10

-4/3

-1

-2/3

-1/3

1/3

2/3

4/3

5/3

With the Table 3 is easy to understand the following equivalence:

(2.2)

Now for the definition (1) of the Volonterios Transform, we can write the following relation (see tables and
definitions attached):
(2.3)

Of course the relation (2.3) is the one that will lead us to obtain the generalized solution of the expression (2.1).

27

We consider the following known relationship [A.V. Oppenheim R.W. Schafer Elaborazione numerica dei segnali (Digital Signal Processing)
Franco Angeli Editions]:
(2.4)

and more in general with NM we define the following expression:

(2.5)

of the (2.4) we must narrow the field to only non-negative integers, namely the (2.4) there must return the zero for
each integer value negative, so we have to rewrite (2.4) as follows:

(2.6)

where

is a step discrete function of Heaviside defined as:

(2.7)

furthermore, we observe that:

(2.8)

Now placing

we deduce the discrete function

(2.9)

and then through the observation (2.8):

(2.10)

Or the equivalent expression:

(2.11)

28

So to utilize the transformed V is useful to rewrite (2.10) in the following way:

(2.12)

then we can rewrite the relation (2.3) as follows:

and thence:

(2.13)

From the PF N. 9 which is given here for convenience:

(2.14)

we have:
(2.15)

While for the PF N.5 which is given here with

(2.16)

and thence:
(2.17)

We obtain finally the following equivalence:

(2.18)

To facilitate the understanding of the examples that follow will call with

29

and with

manipulating the

in the following way:

(2.19)

we observe that the inside bracket of the (2.19) is precisely the function

(2.20)

30

CASE 1 (N < M):

Example 1.1:

where k goes from 0 to 3

-2

-1

-2

thence:
(2.21)

Example 1.2:

with k from 0 to 10:

-3

-2

-1

-1

11

15

thence:

(2.22)

31

Example 1.3:

where k goes from 0 to 9

-6

-5

-4

-3

-2

-1

-2

10

12

14

16

thence:

(2.23)

32

CASE 2 (N = M)
Example 2.1:

where k goes from 0 to 4

-2

-1

-5

10

15

20

thence:

(2.24)

Example 2.2:

where

goes from 0 to 6

-2

-1

-7

14

21

28

thence:

(2.25)

33

CASE 3 (N > M):

Example 3.1:

where k goes from 0 to 4

-2

-1

-9

-2

12

19

26

thence:

(2.26)

***
In general from these observations by induction we conclude for deduction the following equivalence:

(2.27)

Where with
simpler form:

represent the function that returns the integer part of , and then we can rewrite the (2.18) in a

(2.28)

where if

and in particular for M 0 we have

thence:

(2.29)

34

PARTICULAR CASES
Case with
Case with

(2.30)

It can be shown easily from (2.2) and (2.3) that the function

satisfies the following differential equation:

(2.31)

thence:

(2.32)

we have for the TF N.2

, furthermore we have:

(2.33)

thence:

(2.34)

after several steps we obtain:

(2.35)

where here we separate the real part from the imaginary part:
(2.36)

it follows:

35

(2.37)

and thence:

(2.38)

In particular for

we have:

(2.39)

From 2.33 it is shown that:

and

Putting

but

we obtain

and thence

36

Case with
Rewriting the eq. (2.18) as follows:

(2.40)

in this eq. we replacing the following expression

and thence

(2.41)

***
EXAMPLES
Example 1
Calculate
Solution
We consider

with N=2 and M=1 and we replace in the (2.41) obtaining:

(2.42)

the third summation of the (2.42) is null because M<M, thence:

but

then we proceed in the steps:

37

(2.43)

(2.44)

(2.45)

(2.46)

for the TF.5 we have:

(2.47)

(2.48)

Example 2
Calculate
. We shall see how in this example the Volonterios polynomials lend themselves to providing a
generalized solution of this problem.
Solution
In this case is M=0 thus the last term of the (2.30) is null:
(2.49)

For the TF. N.4 we obtain:

38

(2.50)

where

is the Volonterios polynomial of order n.

Putting

and

we have

:
(2.51)

(2.52)

(2.53)

Here we can connect with the Ramanujans equation concerning the number 8, that is a Fibonaccis
number and is linked to the physical vibrations of the superstrings, i.e.

cos txw '

x 2 w '
e
dx
0 cosh x

142
4anti log

2
t
t 2 w'

w'

4
e
w' itw '
1
W02 k 3 , t 1 /
t 3t cosh t 1 t 2 sinh t .
3
10 11 2
10 7 2

log

4
4

(2.53b)

Example 3
Calculate

Solution
(2.54)

(2.55)

39

(2.56)

(2.57)

Now to calculate the transform, we use the TF No. 17:

(2.58)

Thence:
(2.59)

After some calculations we have:

(2.60)

Example 4
Calculate
Solution
We'll see how this example is related to the Laguerres polynomials.
(2.61)

(2.62)

(2.63)

from the TF N. 15 we have:

(2.64)

then replacing:
40

(2.65)

Putting N=2 and M=3 we have

:
(2.66)

(2.67)

(2.68)

(2.69)

(2.70)

This last equation can be connected with the Euler Gamma Function and with the number linked to the physical
vibrations of the bosonic strings, i.e. 24.
Indeed, with regard the number 24, this is related to the modes that correspond to the physical vibrations of the
bosonic strings by the following Ramanujan function:

cos txw '

x 2 w '
0 cosh x e dx 142

4anti log
2
t 2

w'

t w'
4

itw
'
w
'

.
24
10 11 2
10 7 2

log

4
4

Thence, we have the following expression:

41


cos txw '

x 2 w '
e
dx

0 cosh x
142
4anti log
2
t 2

w'

t w'
4

itw
'
w
'
2k 7

W23
, t :
t 18t t 3 cosh t 6 9t 2 sinh t .
10 11 2
2k 4
10 7 2

log

4
4

(2.70b)

ON SOME MATHEMATICAL CONNECTION WITH SOME SECTORS OF STRING THEORY

In 1968 Veneziano proposed the following heuristic answer

As, t

s t
s t (3.1)

s 0 ' s .

with

Euler Gamma function has poles in the negative real axis at integer values

s n

with residue

s 1
s n 1

s
s s 1 s 2 ... s n 1 s n
s n

Hence, at fixed

1n

1
n! s n

(3.2)

t , the amplitude has infinitely many poles at s 0, for s 0 s ' n or

s

n 0
M n2
'
(3.3)

with residue
42

s n

s, t

1n

t 1 t 2... t n 1
t
1

n! n t s n
n!
s n . (3.4)

2
In the bosonic string the simplest vertex operator is the one for the tachyon state N 0 hence M 4 / ' .
We have:

V0; p g s d 2 zeip X g s d 2 zV z , z ; p

. (3.5)

With regard the 4-point tachyon amplitude, we have the following equation:

m
' p j pl

g sm 2
m i , pi pi
d 2 zi z j zl

j l
i
Vol SL2, C i 1
. (3.6)

Setting

m 4 we end up with
4
' p j pl

g s2
i , pi pi
d 2 zi z j zl

j l
i
Vol SL2, C i1
. (3.7)

After fixing the SL2, C invariance by putting the insertion points at 0,1, z and

z4 we end up with

'p p
'p p
4 g s2 pi d 2 z z 1 3 1 z 2 3
i

(3.8)
using Gamma function identities this expression can be given a nice form. One must use the integral representation
2
d z z

where

2a 2

1 z

2b 2

2a b c
1 a 1 b 1 c

(3.9)

a b c 1 . With this, (3.8) can be shown to be equal to

1 ' s / 4 1 ' t / 4 1 ' u / 4

4 g s2 pi
i
2 ' s / 4 2 ' t / 4 2 ' u / 4 (3.10)

in terms of the Mandelstam variables

43

2
s p1 p2 ; t p2 p3 2 ; u p1 p4 2

which satisfy on shell (i.e. use the tachyon mass

(3.11)

pi2 M 2 4 / ' )
4

s t u pi2 M i2
i 1

16
' (3.12)

We can write also the following mathematical connection:

'p p
'p p
4 g s2 pi d 2 z z 1 3 1 z 2 3
i

1 ' s / 4 1 ' t / 4 1 ' u / 4

g s2 pi

i
2 ' s / 4 2 ' t / 4 2 ' u / 4
4

pi2 M i2
i 1

16
'

(3.13)

This expression can be related with the following Ramanujans modular equation linked with the modes (i.e. 8 that
is also a Fibonaccis number) that correspond to the physical vibrations of the superstrings:

cos txw'

x 2 w '
e
dx

142
0 cosh x
4 anti log
2
t 2
t w'

w'

e 4 w' itw'
1
8
3
10 11 2
10 7 2

log

4
4

. (3.14)

Thence, we have the following relationship:

'p p
'p p
4 g s2 pi d 2 z z 1 3 1 z 2 3
i

1 ' s / 4 1 ' t / 4 1 ' u / 4

g s2 pi

i
2 ' s / 4 2 ' t / 4 2 ' u / 4
4

pi2 M i2
i 1

44

16

'

 cos txw'

x 2 w '
e
dx

142
0 cosh x
4 anti log
2
t 2
t w'

w'

e 4 w' itw'
1
2
3
10 11 2
10 7 2

' log

4
4

. (3.15)

We note that this relationship ca be related also with the eq. (d), i.e. the inverse transform of
obtain this further mathematical connection:

V t , thence we

T V t , k : k 1 R V ei cosk d

'p p
'p p
g s2 pi d 2 z z 1 3 1 z 2 3
i

1 ' s / 4 1 ' t / 4 1 ' u / 4

g s2 pi

i
2 ' s / 4 2 ' t / 4 2 ' u / 4
4

pi2 M i2
i 1

16

'

cos txw'

x 2 w '
0 cosh x e dx 142

4 anti log
2
t 2
t w'

w'

4
e
w' itw'
1
2
3
10 11 2
10 7 2

' log

4
4

. (3.15b)

b)the open string scattering

With regard the open string scattering, the amplitude is computed with operator insertions along the boundary of
the disk which maps onto the real axis of the complex plane. The equation of the amplitude is:

4
2 ' p i p j
4
gs
gs
26
ip1 X x1
ip1 X x 4

p
dx
x

dx
e
...
e

i
i
i
j

i
Vol SL2, R i
Vol SL2, R i 1
j l
i 1
.

(3.16)
For a given ordering, the residual symmetry can be used to fix 3 points to
resulting expression contains a single integration for 0 x 1
45

x1 0, x2 0, x3 x

and x4 . The

4 g s dx x
1

2 ' p1 p 2

1 x

2 ' p 2 p 3

. (3.17)

This integral is related to the Euler Beta function (thence with the Euler Gamma function)

Ba, b dxx a 1 1 x
1

b 1

a b
a b . (3.18)

Whence, using now the tachyon mass M 1 / ' one recovers the Veneziano amplitude
2

1 ' s 1 ' t
4 g s

1 ' s t . (3.19)

Thence, we have the following possible mathematical relationship between 3.16), (3.17) and (3.19):
4

4
2 ' p i p j
4
gs
26
gs
ip1 X x1
ip1 X x 4

p
dx
x

dx
e
...
e

i
i
i
j

i
Vol SL2, R i
Vol SL2, R i 1
j l
i 1

g s dx x

2 ' p1 p 2

2 ' p 2 p 3

1 x

1 ' s 1 ' t

1 ' s t . (3.20)

g s

Also this relationship can be related with eq. (d), thence we obtain this further mathematical connection:
T 1 V t , k :

k 1 R V ei cosk d

4
2 ' p p
4
gs
gs
ip1 X x1
ip1 X x 4
26

dx
e
...
e

i dxi xi x j

Vol SL2, R i
Vol SL2, R i 1
j l
i 1
i

g s dx x
0

2 ' p1 p 2

1 x

2 ' p 2 p 3

1 ' s 1 ' t
g
s
1 ' s t

. (3.20b)

c) Four point amplitude for the tachyons from CFT

The ground state tachyons in the twisted sector corresponds to:

46

M 2 1
N (3.21)
For the near marginal tachyons, in the large N limit, which are in the N k th sector, the vertex operator in the
1,1 representation is,

1, 1

z, z e

k
k ~
~ i 1 H z i 1 H z
N
N

e e

eik . x z , z

. (3.22)

The four point amplitude for these lowest lying tachyons can now be computed by taking two vertices in the

0,0 representation and two in the 1,1 representation.

~~
~~
C d 2 z V1, 1 z , z eTF e TFV1, 1 1V1, 1 z , z eTF e TFV1, 1 0
C

The constant

C g c4Cs2

, where

Cs2

is related to

gc

. (3.23)

by

Cs2

4
g c2 . (3.24)

This amplitude can now be computed and is given by,

I C k1.k3

d 2z

2 s

1 z

F z

2 t

, (3.25)

where F z is the hypergeometric function,

k
k
k
k
1 1
1

F z F ,1 ;1; z dyy N 1 y N 1 yz N
N
N
0
, (3.26)
k

and

2
2
2
s k1 k2 , t k2 k3 , s k3 k1 .

47

In the large N approximation,

F z 1

k
1 2 1 3

2
z z z ... O k / N
N
2
3

. (3.27)

2
Note that the terms proportional to k / N in (3.27) shift the s-channel pole. There is an additional factor of k1.k2 ,

2
due to which the contact term from any of the terms of (3.27) apart from 1, would at least be of Ok / N . With this

observation, the integral can now be performed for F z 1 .

s t
s t
1
2 2 2 2
2
I C 2 k1.k3

s
t
s t
1 1
2 2 2 2
s
t
s

1 1 1
2 1
1
1
2 2 2
2
4 g c2 u 2m 2
4
s t 1 s t 1 s 1

2 2 2

Now using

2
t

. (3.28)

s t u 4m 2 ,

t 2m 2
I 4 2 g c2
s

s 2m
2

2 2

s
t
s

1 1 1

2 2 2
3s t 8m 2
s t
s
1 1 1
2 2 2

where we have to expand the gamma functions.

Also here we can write the following relationship between (3.25) and (3.29):

I C k1.k3

dz
2

48

2 s

1 z

F z

2 t

2
t

2 . (3.29)

t 2m 2
4 2 g c2
s

s 2m
2

2 2

t
s t
s
1 1 1

2 2 2 2
3s t 8m 2
1 s t 1 s 1 t
2 2 2 2 (3.30)

Also this expression can be related with the eq. (d) and with the Ramanujans modular equation concerning the
number 8 and thence, we obtain this further mathematical connection:

T 1 V t , k :

k 1 R V ei cosk d

C k1.k3

t 2m
4 2 g c2
s

s 2m

2 2

2 2

d
C

2 s

1 z

F z

2 t

t
s t
s

2 2 2 2
2

3s t 8m

1 s t 1 s 1 t
2 2 2 2

cos txw'

x 2 w '
0 cosh x e dx 142

4 anti log
2
t 2
t w'

w'

e 4 w' itw'
1

3
10 11 2
10 7 2

log

4
4

(3.30b)

d) expressions concerning the four tachyon amplitude in CSFT

With regard a closed analytical expression for the off-shell four tachyon amplitude in CSFT, Giddings gave an explicit
conformal map that takes the Riemann surfaces defined by the Witten diagrams to the standard disc with four
tachyon vertex operators on the boundary. This conformal map is defined in terms of four parameters , , , .
The four parameters are not independent variables. They satisfy the relations

1 , 1 (3.31)

and

1
0 1 , k 0 2 , k
2
, (3.32)

where

0 , k

is defined by
49

 0 , k

E k F , k ' K k E , k ' K k F , k '

. (3.33)

In (3.33) K k and E k are complete elliptic functions of the first and second kinds, F , k is the incomplete elliptic
integral of the first kind. The parameters 1 , 2 , k and k ' satisfy

k2

sin 1

2
2

k '2 1 k 2 , (3.34)

sin 2

2 2 ,

2 2 . (3.35)

By using the integral representations of the elliptic functions it is possible to write the equation (3.32) in a useful
form

E 2

dt

To expand (3.36) for small

t2 4 1 t2

1 4 K 2

dt

t2 4 1 t2

4
. (3.36)

and we have to divide the integration region into three intervals in such a way that

the square roots in the denominators of (3.36) can be consistently expanded and the integrals in
example consider the integral in the first term of (3.36) , it can be rewritten as

dt

1
t
2

1 t

dt

2 1

1 t2

2 dt

t 1

1 t2

t performed. For

dt
t2 1

4
t

1
t 2 (3.37)

In each integral of the rhs the integration domain is contained in the convergence radius of the Taylor expansions of

the square roots containing , so that they can be safely expanded and the integrals in
procedure one gets the following equation equivalent to (3.36):

t performed. With this

4 k 2 n 1

2
2
2k
2

1 kn

2n 2k 1
1

n, k 0 1

n k n!k!
2
2

50

1 1

4 k 2 n 1

1
2k
2 2

2n 2k 1
1

n, k 0 1

n k n!k!
2
2

kn 4 n ln 2 1 4 K 2
2k 2n

4n

4k

1 kn

4n 4k
2k 2n

kn

4n

4n 2k 3

1
2n
ln

2n 2k 3

4
2

(3.38)
Thence, from (3.36) and (3.38) we can write the following mathematical relationship:

E 2

dt

1
t
2

1 t

1 4 K 2

dt

1
t
2

1 t
2

4 k 2 n 1

2
2
2k
2

E
1 kn

2n 2k 1
1

n, k 0 1

n k n!k!
2
2

1 1

4 k 2 n 1

1
2 2
2k
4n
2
4
2

kn ln 1 K

2k 2n
2n 2k 1
1

n, k 0 1

n k n!k!
2
2

4n

4k

1 kn

4n 4k
2k 2n

kn

4n

4n 2k 3

1
2n
ln

2n 2k 3

4
2

. (3.39)

Also this expression can be related with the eq. (d), and thence we obtain this further mathematical connection:

T V t , k : k 1 R V ei cosk d

E 2

dt

1
t
2

1 t

1 4 K 2

51

dt

1
t
2

1 t
2

4 k 2 n 1

2
2
2k
2

1 kn

2n 2k 1
1

n, k 0 1

n k n!k!
2
2

kn 4 n ln 2 1 4 K 2
2k 2n

4n 4k

1 kn

4n 4k
2k 2n

1 1

4 k 2 n 1

1
2k
2 2

2n 2k 1
1

n, k 0 1

n k n!k!
2
2

kn

4n

4n 2k 3

1
2n
ln

2n 2k 3

4
2

e) physical interpretation of the nontrivial zeta zeros in terms of tachyonic string poles

The four-point dual string amplitude obtained by Veneziano was

A4 As, t At , s Au , s dx x

1 x

B ,

, (3.41)

where the Regge trajectories in the respective s, t , u channels are:

s 1

1
1
1
s t 1 t u 1 u
2 ,
2 ,
2 . (3.42)

The conservation of the energy-momentum yields:

k1 k2 k3 k4 k1 k2 k3 k4 0

. (3.43)

We have also that the sum

s t u 2 k12 k22 k32 2k1 k2 k2 k3 k1 k3 8

(3.44)

in mass units of m Planck = 1, when all the four particles are tachyons and one has the on-shell condition:

52

(3.40)

k12 k22 k32 m 2

2m2Planck = 2

(3.45)

in the natural units LPlanck = 1 such that the string slope parameter in those units is given by L2Planck = 1/2 and the
string mass spectrum is quantized in multiplies of the Planck mass mPlanck = 1.
From the conservation of energy-momentum (3.43) and the tachyon on-shell condition eq. (3.45) one can deduce
that:

k1 k2 2 k3 k4 2 k1 k2 k3 k4 . (3.46)
Therefore, from eqs. (3.44) (3.46) it is straightforward to show:

s t u 2 2 2 2 2k1 k2 k3 k1 k2 12 2k1 k2 k3 k3 k4

12 2k1 k2 k3 k4 k3 k3 12 2k3 k3 12 4 8

(3.47)

2
This relationship among s t u 4m 8 will be crucial in what follows next. From eqs. (3.42), (3.44), and (3.47)
we learn that:

1 . (3.48)

There exists a well-known relation among the

functions (Euler Gamma function) in terms of

(Riemann zeta function) appearing in the expression for

case, the integration region in the real line that defines
leads to the very important identity

As, t , u B ,

where

functions

As, t , u when , fall inside the critical strip. In this

As, t , u in eq. (3.41) can be divided into three parts and

1 1 1

(3.49)

1 and , are confined to the interior of the critical strip.

The derivation behind eq. (3.49) relies on the condition 1 eq. (3.48) and the identities

53

sin sin sin 4 cos

cos

cos

2 , (3.50)

sin , (3.51)

plus the remaining cyclic permutations from which one can infer

sin

, (3.52)

sin

, (3.53)

sin

. (3.54)
Therefore, eqs. (3.50) (3.54) allow us to recast the left hand side of (3.49) as

As, t , u B ,

cos

cos

cos

. (3.55)

And, finally, the known functional relation

2 z 1 z 2 cos z z z
2

, (3.56)

2 is what establishes the important identity
in conjunction with the condition 1 such that 2

(3.49) expressing explicitly the string amplitude As, t , u either in terms of zeta functions or in terms of functions.
In conclusion, we have the following interesting relationship between the eqs. (3.41), (3.49) and (3.55):

A4 As, t At , s Au , s dx x
R

54

1 x

B ,

 As, t , u B ,

1 1 1

cos

cos

cos

, (3.57)

from which we can to obtain the following equivalent expression:

1
1
1
1
1
1

A4 dx x 1 x
B , cos cos cos
4
4 R
4

8
8
8
(3.58)

In this expression there are both and 8, i.e. the number that is connected with the modes that correspond to the
physical vibrations of a superstring by the following Ramanujan function:

cos txw'

x 2 w '
e
dx

142
0 cosh x
4 anti log
2
t 2
t w'

w'

e 4 w' itw'
1
8
3
10 11 2
10 7 2

log

4
4

. (3.59)

Thence the final mathematical connection:

1
1
1
1
1
1

A4 dx x 1 x
B , cos
cos
cos
4
4 R
4

8
8
8
cos txw'

x 2 w '
e
dx
0 cosh x

142
4 anti log

2
t
t
w'

w'

itw
'
w
'
1

3
10 11 2
10 7 2

log

4
4

(3.60)

55

References

Odoardo Volonterio, Michele Nardelli, Francesco Di Noto On a new mathematical application concerning

the discrete and the analytic functions. Mathematical connections with some sectors of Number
Theory and String Theory. Feb- 2014 http://empslocal.ex.ac.uk/people/staff/mrwatkin//zeta/nardelli2014.pdf

56