1 Page

Equations of Orbits, Deflection of Light,

Mercurys Perihelion Shift, Period of Revolution and
Time Delay of Signals
in Schwarzchilds Geometry.
Some Exact Formulas.
Solomon M. Antoniou
SKEMSYS
Scientific Knowledge Engineering
and Management Systems
37 oliatsou Street, Corinthos 20100, Greece
[email protected]
2 Page
Abstract
We present some exact solutions of the differential equations of motion connected
to Schwarzchilds solution of Einstein field equations in General Relativity. The
exact solutions are expressed in terms of Jacobis or Weirstrasss Elliptic
Functions or in terms of Elliptic Integrals.
The topics which are covered are the following:
I) Deflection of light in the Suns gravitational field
II) Equation of a closed orbit
III) Mercurys perihelion shift
IV) Trajectory of a light signal
V) Period of revolution
VI) Time delay of radio signals in the Suns gravitational field
Although the subject itself is rather old, some of the results of this paper appear for
the first time in the literature. In particular the formulae (10.29) for the period and
(11.19) for the time delay, to the best of our knowledge, appear for the first time in
the literature.
Keywords: General Relativity, Einsteins Equations, Schwarzchild Solution,
Equations of Orbits, Mercurys Perihelion Shift, Deflection of Light, Period of
Revolution, Time Delay of Signals, Weirstrass Elliptic Functions, Jacobi Elliptic
Functions, Elliptic Integrals, Exact Solutions.
3 Page
Contents
1. In t r odu ct ion
2. Th e Sch wa r zch ild Solu t ion
3. Equ a t ion s of t h e Geodes ic Cu r ves
4. Simplifica t ion of t h e Equ a t ion s of Mot ion
5. Differ en t ia l Equ a t ion s . Ellipt ic Fu n ct ion s
5.1 Differ en t ia l Equ a t ion of t h e Or bit
5.2 Differ en t ia l Equ a t ion for t h e Ca lcu la t ion of t h e Time
6. Deflect ion of Ligh t in t h e Su n s Gr a vit a t ion a l Field
7. Equ a t ion of Clos ed Or bit
7.1 Equ a t ion of Clos ed Or bit u s in g Weir s t r a s s Ellipt ic
Fu n ct ion
7.2 Equ a t ion of Clos ed Or bit u s in g J a cobis Ellipt ic
Fu n ct ion
7.3 Equ a t ion of Clos ed Or bit . Th ir d Met h od
8. Mer cu r ys Per ih elion Sh ift
9. Tr a ject or y of Ligh t Sign a l
9.1 Tr a ject or y of Ligh t Sign a l u s in g Weir s t r a s s Ellipt ic
Fu n ct ion
9.2 Tr a ject or y of Ligh t Sign a l u s in g J a cobis Ellipt ic
Fu n ct ion
10. Per iod of Revolu t ion
11. Time Dela y of Ra dio Sign a ls in t h e Su n s Gr a vit a t ion a l Field
4 Page
1. Introduction
We analyze the problem of motion of test particles in the gravitational field
created by a heavy spherically symmetric object of mass M, which produces a
Schwarzchild geometry. We study different types of orbits, which mainly fall into
two categories: bound and unbound. We also consider the period of a test body
describing a cycle around a heavy object (say the Sun) and the time delay of radio
signals in the Suns gravitational field.
In all the cases we express the final solutions in terms of Elliptic Functions
(Jacobis or Weirstrasss) or Elliptic Integrals. We use a number of different
approaches, which are then proved to lead to equivalent solutions.
2. The Schwarzchild Solution.
Einsteins field equations in empty space are (Ref. [1])
0 R
v

where
v
R is Riccis tensor.
By solving Einsteins equations for the general static isotropic metric

2 2 2 2 2 2 2 2
d 0 sin r 0 d r dr ) r ( A dt ) r ( B t d (2.1)
we can determine the functions ) r ( B and ) r ( A .
These functions can be determined by imposing the boundary condition that for
r the metric tensor must approach the Minkowski tensor in spherical
coordinates:
1 ) r ( B lim ) r ( A lim
r r


A constant of integration which appears in the solution, can be fixed considering
that at great distances from a central mass M, the component B g
tt
must
approach to the quantity 2 1 , where is the Newtonian potential
r
G M

5 Page
where M is the mass of the object producing the field and G is Newtons
constant. Under these circumstances, we find that the functions ) r ( B and ) r ( A are
given by

r
MG 2
1 ) r ( B (2.2)
and

1
r
MG 2
1 ) r ( A

,
_

(2.3)
respectively.
The solution of Einsteins equations for the metric (2.1) with ) r ( B and ) r ( A given
by (2.2) and (2.3) respectively, is the very-well known Schwarzchild solution.
It is obvious that
1 ) r ( B ) r ( A (2.4)
The non-vanishing Christoffel symbols associated to the metric (2.1) are given by

) r ( A 2
) r ( A
I

rr


) r ( A
r
I
00
r

) r ( A
0 sin r
I
2

r


) r ( A 2
) r ( B
I
tt
r


r
1
I I
r 0
0
0 r
0
0 cos 0 sin I

0
(2.5)

r
1
I I
r

0 cot I I
0

) r ( B 2
) r ( B
I I
rt
t
tr
t


where the prime denotes differentiation with respect to r.
3. Equations of the Geodesic Curves
The differential equations of the geodesic curves in a curved space have the form
0
dp
dx
dp
dx
I
dp
x d
v
v

2
2
+ (3.1)
where p is an affine parameter and summation over repeated indices is always
understood.
6 Page
Using the non-vanishing components of the Christoffel symbols (relations (2.5)),
we find from equation (3.1) the equations
+

,
_

,
_

,
_

+
2
2
2 2
2
2
dp
d
) r ( A
0 sin
r
dp
0 d
) r ( A
r
dp
dr
) r ( A 2
) r ( A
dp
r d
0
dp
dt
) r ( A 2
) r ( B
2

,
_

+ (3.2)
0
dp
d
0 cos 0 sin
dp
dr
dp
0 d
r
2
dp
0 d
2
2
2

,
_

+ (3.3)
0
dp
0 d
dp
d
0 cot 2
dp
dr
dp
d
r
2
dp
d
2
2
+ + (3.4)
0
dp
dr
dp
dt
) r ( B
) r ( B
dp
t d
2
2

+ (3.5)
Since the field is isotropic, we may consider the orbit of our particle to be confined
to the equatorial plane, that is

2
a
0 (3.6)
Equation (3.3) is then immediately satisfied and we can forget about 0 as a
dynamical variable.
Under this assumption, the system of equations (3.2)-(3.5) is equivalent to the
system
0
dp
dt
) r ( A 2
) r ( B
dp
d
) r ( A
r
dp
dr
) r ( A 2
) r ( A
dp
r d
2 2 2
2
2

,
_

,
_

,
_

+ (3.7)
0
dp
dr
dp
d
r
2
dp
d
2
2
+ (3.8)
0
dp
dr
dp
dt
) r ( B
) r ( B
dp
t d
2
2

+ (3.9)
7 Page
4. Simplification of the Equations of Motion
We can further simplify the equations of motion (3.7)-(3.9) we have derived. Full
details are given in Appendix A. The new equations are the following:

) r ( B
1
dp
dt
(4.1)
J
dp
d
r
2
(4.2)
E
) r ( B
1
r
J
d
dr
) r ( A
2
2
2
+

,
_

(4.3)
where J is a constant having dimensions of angular momentum per unit mass and
E is a constant.
The relation between proper time t and affine parameter p is given by

2 2
dp E t d (4.4)
where
0 E > for particles
0 E for photons
The equation of the orbit ) ( r r is obtained by eliminating dp between
equations (4.2) and (4.3). We find

2 2 2
2
4
J
E
) r ( B J
1
r
1
d
dr
r
) r ( A
+

,
_

(4.5)
which can be put in one of the following equivalent forms:

,
_

,
_

2 2 2
4
2
r
1
J
E
) r ( B J
1
) r ( A
r
d
dr
(4.6)
and
1
r
J
E
) r ( A
1
d
dr
r
J
2
2
2
2

,
_

+ +

,
_

(4.7)
8 Page
We can also determine the value of the constant J as follows:
If the orbit is an unbounded curve, at the closest distance from the object which
produces the gravitational field, we have
0
r r and so 0
d
dr
, and equation (4.5)
gives

,
_

E
) r ( B
1
r J
0
2
0
2
(4.8)
In case the path is a closed curve, we find another expression for J (Appendix D):

) r r )( MG 2 r )( MG 2 r (
) r r ( MG 2
J
2
2
+ +
+
+

(4.9)
where
+
r and

r are the aphelion and perihelion respectively.

Finally, since dt ) r ( B dp , as follows from equation (4.1), we can eliminate dp
from equation (4.3) and we get
E
) r ( B
1
r
J
dt
dr
) r ( B
) r ( A
2
2
2
2
+

,
_

(4.10)
5. Differential Equations. Elliptic Functions
5.1. Differential Equation of the Orbit.
The differential equation of the orbit can be expressed in the form
) r ( f
d
dr
2

,
_

(5.1)
where ) r ( f is a fourth degree polynomial, which can easily be integrated.
Under the substitution
r
1
u , equation (5.1) is being converted into an equation of
the form
) u ( g
d
du
2

,
_

(5.2)
9 Page
where ) u ( g is a third degree polynomial. The resulting equation can then be
integrated. The resulting integral is then converted to an elliptic integral or to a
linear combination of elliptic integrals of various types.
We can also transform the differential equation (5.2) into a Weirstrass or Jacobi
differential equation. In fact, under a linear transformation U u + , the
differential equation is converted to

3 2
3
2
g U g U 4
d
dU

,
_

(5.3)
The above differential equation has the general solution
) g , g ; c ( U
3 2
+ (5.4)
where is Weirstrass elliptic function and c is a constant which can be
determined from the initial conditions.
Since, on the other hand,
) e U )( e U )( e U ( 4
d
dU
3 2 1
2

,
_

(5.5)
where
3 2 1
e e e > > , under the substitution

3 2
3 2
e e
e U
z

,
3 1
3 2 2
e e
e e
k

( 1 k 0
2
< < ) (5.6)
the differential equation is transformed to
) z k 1 )( z 1 (
dx
dz
2 2 2
2

,
_

(5.7)
where e e x
3 1
and it has the general solution
) k ; c x ( sn z + (5.8)
where ) x ( sn is Jacobis elliptic function.
10 Page
5.2. Differential equation for the calculation of time.
The differential equation for the calculation of time in terms of the distance is
given by (Appendix A, equation (A.19))
) r ( R
dr
dt
(5.9)
where

2 / 1 3 2 3
)] x ( B x E ) x ( B x J x [
x x
MG 2 x
x
) x ( R

(5.10)
The above differential equation can be integrated by separation of variables and its
solution can be expressed as a combination of Elliptic Integrals of various types.
6. Deflection of Light in the Suns Gravitational Field.
We start with equation (4.6) where
2
J is given by equation (A.11) of Appendix A.
Since

2 / 1
2 2 2
2
r
1
J
E
) r ( B J
1
) r ( A
r
d
dr

,
_

t
we obtain, by separation of variables
dr
r
1
J
E
) r ( B J
1
r
) r ( A
d
2 / 1
2 2 2
2

,
_

t (6.1)
We also have

,
_

2 2 2 2 2
r
1
E
) r ( B
1
J
1
r
1
J
E
) r ( B J
1

2
1
0
2
0
r
1
E
) r ( B
1
E
) r ( B
1
r
1

,
_

,
_

(6.2)
and thus equation (6.1) can be written as
11 Page
dr
r
1
E
) r ( B
1
E
) r ( B
1
r
1
r
) r ( A
d
2 / 1
2
1
0
2
0
2
1
1
1
]
1

,
_

,
_

(6.3)
Integrating the previous equation, we obtain the equation of the orbit

1
1
1
]
1

,
_

,
_

+
r
2 / 1
2
1
0
2
0
2
dr
r
1
E
) r ( B
1
E
) r ( B
1
r
1
r
) r ( A
) ( ) r ( (6.4)
This is the general equation of the orbit of a particle, which approaches the
gravitational field produced by an object of mass M.
If the particle is a photon, then 0 E and equation (6.4) takes the form

1
1
]
1

r
2 / 1
2 2
0
0
2
dr
r
1
) r ( B r
) r ( B
r
) r ( A
) ( ) r (

1
1
]
1

r
2 / 1
2 2
0
2
0 0
2
2
dr
) r ( B r r
) r ( B r ) r ( B r
r
) r ( A
) (

+
r
2 / 1 2
0 0
2
0
dr
)] r ( B r ) r ( B r [ r
) r ( B ) r ( A r
) (
and since 1 ) r ( B ) r ( A , we get finally

r
2 / 1 2
0 0
2
0
dx
)] x ( B r ) r ( B x [ x
1
r ) ( ) r ( (6.5)
The integral which appears in (6.5) can be transformed as follows. First of all we
have to transform the quantity under the integration. In Appendix B we find that
12 Page

2 / 1 2
0 0
2
)] x ( B r ) r ( B x [ x
) r x )( r x )( r x ( x
r
MG 2 r
2 1 0
0
0

(6.6)
where

,
_

+
+
MG 2 r
MG 6 r
1 r
2
1
r
0
0
0 1
and

,
_

+

MG 2 r
MG 6 r
1 r
2
1
r
0
0
0 2
(6.7)
Collecting everything together, we have

r
2 1 0 0
0
0
dx
) r x )( r x )( r x ( x
1
MG 2 r
r
r ) ( ) r ( (6.8)
We put
) r x )( r x )( r x ( x ) x ( G
2 1 0
(6.9)
We have to calculate the difference

0
r
0
0
0 0
dx
) x ( G
1
MG 2 r
r
r ) ( ) r ( (6.10)
In Appendix C it is proved that

a
0
2 2
1 2 0
r
sin k 1
d
) r r ( r
2
dx
) x ( G
1
0
where

,
_

1 0
1 2 1
r r
r r
sin a and
1 2
1 0
0
2 2
r r
r r
r
r
k

Equation (6.10) is then equivalent to

a
0
2 2
1 2 0 0
0
0 0
sin k 1
d
) r r ( r
2
MG 2 r
r
r ) ( ) r (
13 Page
) k ; a ( F
) r r )( MG 2 r (
r 2
1 2 0
0

(6.11)
where ) k ; a ( F is the Elliptic Integral of the first kind.
We also have

MG 2 r
MG 6 r
r r r
0
0
0 1 2

+
(6.12)
and thus equation (6.11) is equivalent to
) k ; a ( F
) r r )( MG 2 r (
r
2 ) ( ) r (
4 / 1
1 2 0
2
0
0
1
1
]
1

(6.13)
The deflection A of the light is given by the equation
a | ) ( ) r ( | 2 A
0
(6.14)
and using (6.13), we get the expression
a ) k ; a ( F
) r r )( MG 2 r (
r
4 A
4 / 1
1 2 0
2
0

1
1
]
1

(6.15)
7. Equation of Closed Orbits.
7.1 Equation of Closed Orbit using Weirstrass Elliptic Function
We start with equation (4.6):

) r ( A
r
) r ( A
r
J
E
r
J
1
d
dr
2 4
2
4
2
2

,
_

(7.1)
which, using (7.30), (7.31) and the expression for ) r ( A , can be converted to the
equation
r MG 2 r r
b
G M a 2
r
b
a c
d
dr
2 3 4
2
+

,
_

(7.2)
The quantities a, b and c are defined by (7.27), (7.28) and (7.29) respectively.
Under the substitution
14 Page

r
1
u ,
d
du
u
1
d
dr
2
(7.3)
equation (7.2) can be transformed to the equation

3 2
2
u MG 2 u u
b
G M a 2
b
a c
d
du
+ +

,
_

(7.4)
We try a linear transformation
U u + (7.5)
so as to convert (7.4) into the standard form:

3 2
3
2
g U g U 4
d
dU

,
_

(7.6)
Under (7.5), equation (7.4) is transformed to

3 2
2
2
) U ( MG 2 ) U ( ) U (
b
G M a 2
b
a c
d
dU
+ + + + +

,
_

which is equivalent to
+

,
_

+ + + +

,
_

U MG 6 2
b
aMG 2

1
U ) MG 6 1 ( U MG 2
d
dU
2 2 3
2

,
_

+ +

+ MG 2
b
MG a 2
b
a c

1
3 2
2
(7.7)
The coefficients and are to be determined so as the coefficient of
3
U
should be 4 and the coefficient of
2
U should be zero:
4 MG 2 and 0 MG 6 1 +
Solving the previous system we find the values of the constants and :

MG
2
(7.8)
and

MG 6
1
(7.9)
15 Page
Thus the linear transformation (7.5) has the form

MG 6
1
U
MG
2
u + (7.10)
After substituting the values of and into (7.7) we arrive at the equation

3 2
3
2
g U g U 4
d
dU

,
_

(7.11)
where

b
G M a
12
1
g
2 2
2
(7.12)
and

4
G M
b
a c
12
G M
b
a
216
1
g
2 2 2 2
3

(7.13)
The solution to (7.11) is given by
) g , g ; c ( U
3 2
+ (7.14)
where is the Weirstrass elliptic function and c is a constant which can be
determined from the initial conditions.
We thus arrive at the following solution of equation (7.1):

MG 6
1
) g , g ; c (
MG
2
r
1
3 2
+ + (7.15)
We can determine the constant c by imposing the initial condition:
0 ,

r r

MG 6
1
) g , g ; c (
MG
2
r
1
3 2
+

from which there follows:

r 12
r MG 6
) g , g ; c (
3 2
(7.16)
On the other hand, it is a very well-known fact (Ref. [6]) that if
16 Page
) g , g ; z (
3 2

then

3 2
3
g x g x 4
dx
z
which means that

r 12
r MG 6
) g , g ; c (
3 2
implies that

2
e
~
3 2
3
g x g x 4
dx
c
where
2
e
~
is one of the roots of the polynomial
3 2
3
g x g x 4 where
2
g and
3
g
are given by (7.12) and (7.13) respectively.
We also have
) e
~
x )( e
~
x )( e
~
x ( 4 g x g x 4
3 2 1 3 2
3

where

+
+ +
+

r r
) r r ( MG 3 r r
e
~
1
,

r 12
r MG 6
e
~
2
,
+
+

r 12
r MG 6
e
~
3
are the three roots of the polynomial with ordering

3 2 1
e
~
e
~
e
~
> >
Using (17.4.65) of ref. [5], we get
) m ; ( F
e
~
e
~
1
) e
~
x )( e
~
x )( e
~
x ( 4
dx
c
2 1
e
~
3 2 1
2

(7.17)
where

0
2 2
0 sin m 1
0 d
) m ; ( F (7.18)
17 Page
with

3 1
3 2 2
e
~
e
~
e
~
e
~
m

and

,
_

3 2
3 1 1
e
~
e
~
e
~
e
~
sin (7.19)
7.2 Equation of Closed Orbit using Jacobi Elliptic Function.
First Case. Solution of the Equation (7.4):
We have that (7.4) is equivalent to
) e u )( e u )( e u ( MG 2
d
du
3 2 1
2

,
_

where
3 2 1
e , e , e are the three roots of the equation
0 u MG 2 u u
b
G M a 2
b
a c
3 2
+ +

We can find that these roots are given by

h
1
r r MG 2
) r r ( MG 2 r r
e
1

+

+
+ +
,

r
1
e
2
,
+

r
1
e
3

We remind the reader the assumption
MG 2 h r r > > >
+
and MG 6 r >

from which there follows the ordering

3 2 1
e e e > > .
Under the substitution

3 2
3 2
e e
e u
z

3 1
3 2 2
e e
e e
k

( 1 k 0
2
< < )

3 1
2
e e
we get

2
3 2 3
z ) e e ( e u +
) z k 1 ( e u
2 2 2
1

18 Page
) z 1 ( ) e e ( e u
2
3 2 2


2
3 2 3
z ) e e ( e u
dz z ) e e ( 2 du
3 2


d
dz
z ) e e ( 2
d
du
3 2

The lhs of equation (7.4) is transformed into

2
2 2
3 2
2
d
dz
z ) e e ( 4
d
du

,
_

,
_

The rhs of equation (7.4) is transformed into

) e u )( e u )( e u ( MG 2
3 2 1
] z ) e e [( )] z 1 ( ) e e ( [ )] z k 1 ( [ MG 2
2
3 2
2
3 2
2 2 2
) z k 1 ( ) z 1 ( z ) e e ( MG 2
2 2 2 2 2
3 2
2

Therefore the original equation (7.4) is being transformed into the equation
) z k 1 ( ) z 1 ( MG 2
d
dz
4
2 2 2 2
2

,
_

which is equivalent to
) z k 1 ( ) z 1 (
d
dz
2 2 2
2

,
_

or
) z k 1 ( ) z 1 (
dx
dz
2 2 2
2

,
_

(7.4a)
where
x and
2
) e e ( MG

2
MG

3 1


The general solution of the equation (7.4a) is given by
19 Page
) k ; o x ( sn z + , o constant
and so
) k ; o ( sn ) e e ( e u
2
3 2 3
+ +
The equation of the orbit is thus written as

) k ; o ( sn ) e e ( e
1
r
2
3 2 3
+ +

We may calculate the constant o in the previous formula by imposing some

initial conditions:
i) If we impose 0 for
+
r r then from the equation of the orbit we have

) k ; o ( sn ) e e ( e
1
e
1
r
2
3 2 3
3 +

+
from which we get
0 ) k ; o ( sn
2

and then
0 o
In this case the equation of the orbit is

) k ; ( sn ) e e ( e
1
r
2
3 2 3
+

We can now transform this equation in a more standard form.

It is known that if ) k ; a ( F then a sin ) k ; ( sn . We then have

2
a 2 cos 1
a sin ) k ; ( sn
2 2


Introducing the angle 0 by | a 2 a | 0 , we find a 2 cos 0 cos and so

2
0 cos 1
) k ; ( sn
2
+

We obtain in this way

20 Page

2
0 cos 1
) e e ( e
1
r
3 2 3
+
+

from which there follows that

0 cos e 1
R
r
+

where

3 2
e e
2
R
+
and
3 2
3 2
e e
e e
e
+

Since

+
+

r
1
r
1
2
e e
2
R
3 2
and

,
_

+
+
r
1
r
1
2
1
R
1
,
R may be identified with the semi-latus rectum.
We also have

+
+
+

r r
r r
e e
e e
e
3 2
3 2
If a is the semi-major axis, then
a ) e 1 ( r +
+
and a ) e 1 ( r

ii) If we impose 0 for

r r , then from the equation of the orbit we have

) k ; o ( sn ) e e ( e
1
e
1
r
2
3 2 3
2 +

from which we get

2
a
sin 1 ) k ; o ( sn 1 ) k ; o ( sn
2

and then
) k ( K ) k , 2 / a ( F o
In this case the equation of the orbit is
21 Page

) k ; K ( sn ) e e ( e
1
r
2
3 2 3
+ +

Since

) u ( dn
) u ( cn
) k ; ) k ( K u ( sn +
the previous equation of the orbit can be expressed as

) k ; ( dn
) k ; ( cn
) e e ( e
1
r
2
2
3 2 3
+

or in equivalent form

) k ; ( cn ) e e ( ) k ; ( dn e
) k ; ( dn
r
2
3 2
2
3
2
+

Second Case: Solution of the equation (7.11)

We have that (7.11) is equivalent to
) e
~
U )( e
~
U )( e
~
U ( 4
d
dU
3 2 1
2

,
_

where
3 2 1
e
~
, e
~
, e
~
are the three roots of the equation 0 g U g u 4
3 2
3

We can find that these roots are given by

+
+ +
+

r r
) r r ( MG 3 r r
e
~
1
,

r 12
r MG 6
e
~
2
,
+
+

r 12
r MG 6
e
~
3

with the ordering
3 2 1
e
~
e
~
e
~
> > .
Under the substitution

3 2
3 2
e
~
e
~
e
~
U
z

3 1
3 2 2
e
~
e
~
e
~
e
~
k

( 1 k 0
2
< < )

3 1
2
e
~
e
~

22 Page
we get

2
3 2 3
z ) e
~
e
~
( e
~
U +
) z k 1 ( e
~
U
2 2 2
1

) z 1 ( ) e
~
e
~
( e
~
U
2
3 2 2


2
3 2 3
z ) e
~
e
~
( e
~
U
dz z ) e
~
e
~
( 2 dU
3 2


d
dz
z ) e
~
e
~
( 2
d
dU
3 2

The lhs of equation (7.11) is transformed into

2
2 2
3 2
2
d
dz
z ) e
~
e
~
( 4
d
dU

,
_

,
_

The rhs of equation (7.11) is transformed into

) e
~
U )( e
~
U )( e
~
U ( 4
3 2 1
] z ) e
~
e
~
[( )] z 1 ( ) e
~
e
~
( [ )] z k 1 ( [ 4
2
3 2
2
3 2
2 2 2
) z k 1 ( ) z 1 ( z ) e
~
e
~
( 4
2 2 2 2 2
3 2
2

Therefore the original equation (7.11) is being transformed into the equation
) z k 1 ( ) z 1 (
d
dz
2 2 2 2
2

,
_

which is equivalent to
) z k 1 ( ) z 1 (
d
dz
2 2 2
2

,
_

or
) z k 1 ( ) z 1 (
dx
dz
2 2 2
2

,
_

(7.11a)
where
23 Page
x and
3 1
e
~
e
~

The general solution of the equation (7.11a) is given by
) k ; o x ( sn z + , o constant
and then
) k ; o ( sn ) e
~
e
~
( e
~
U
2
3 2 3
+ +
Using (7.10), we arrive at the following equation of the orbit:

MG 6
1
)] k ; o ( sn ) e
~
e
~
( e
~
[
MG
2
r
1
2
3 2 3
+ + +
7.3 Equation of Closed Orbit. Third Method.
Since the path is a closed curve, there are only two points at which 0
d
dr
.
Let us call
t
r these two points. Using equation (4.5) in the form
0 E
) r ( B
1
r
J
d
dr
r
) r ( A J
2
2
2
4
2
+ +

,
_

we get
0 E
) r ( B
1
) r (
J
2
2
+
t
t
(7.20)
In other words
0 E
) r ( B
1
) r (
J
2
2
+
+
+
(7.21a)
0 E
) r ( B
1
) r (
J
2
2
+

(7.21b)
Equations (7.21) is a system of two equations with two unknowns:
2
J and E.
This system is solved in Appendix D. We find

) r r )( MG 2 r )( MG 2 r (
) r r ( MG 2
J
2
2
+ +
+
+
(7.22)
24 Page
and

) r r )( MG 2 r )( MG 2 r (
r r MG 2 ) MG 2 r ( ) r ( ) MG 2 r ( ) r (
E
2 2
+ +
+ + +
+
+
(7.23)
The angle ) r ( through which the radius vector of the particle rotates can be
calculated using (6.1), by the formula

,
_

r
r
2 / 1
2 2 2
2
dr
r
1
J
E
) r ( B J
1
r
) r ( A
) r ( ) r ( (7.24)
With
+
r r we find

,
_

+
r
r
2 / 1
2 2 2
2
dr
r
1
J
E
) r ( B J
1
r
) r ( A
) r ( ) r ( (7.25)
We now have simplify the expression under the integral sign of (7.24). We have

,
_

,
_

2 / 1
2
2
2
2
2 / 1
2 2 2
2
1
J
x E
) x ( B J
x
x
) x ( A
x
1
J
E
) x ( B J
1
x
) x ( A

2 / 1 2 2 4 4
) x ) x ( B J x ) x ( B E x (
) x ( B ) x ( A J

2 / 1 2 3 3
) x ) x ( B J x ) x ( B E x ( x
J

(7.26)
Using the substitutions
) r r ( MG 2 ) MG 2 r ( ) r ( ) MG 2 r ( ) r ( : a
2 2
+ + +
+ (7.27)

2
) r r ( MG 2 : b
+
(7.28)
) r r )( MG 2 r )( MG 2 r ( : c
+ +
+ (7.29)
expressions (7.22) and (7.23) become
25 Page

c
b
J
2
(7.30)
and

c
a
E (7.31)
respectively.
The expression x ) x ( B J x ) x ( B E x
2 3 3
which appears in (7.26) can be further
transformed in Appendix E. We find
x ) x ( B J x ) x ( B E x
2 3 3
)} h x )( r x )( r x ( {
) MG 2 r )( MG 2 r )( r r (
) r MG 2 r MG 2 r r ( MG 2

+

+
+ +
+ +
(7.32)
where

) r r ( MG 2 r r
) r r ( MG 2
h
+ +
+
+
(7.33)
Using now (7.26) and (7.32), we get

,
_

2 / 1
2 2 2
2
x
1
J
E
) x ( B J
1
x
) x ( A

) h x )( r x )( r x (
1
r MG 2 r MG 2 r r
) r r (

+ + +
+
(7.34)
Therefore equation (7.24) takes on the final form


) r ( ) r (

+ + +
+
r
r
) h x )( r x )( r x (
dx
r MG 2 r MG 2 r r
) r r (
(7.35)
The conversion of the integral in (7.35) into an elliptic integral is performed in
Appendix F. We have found that

a
0
2 2
sin m 1
d
) (# T ) r ( ) r ( (7.36)
where
26 Page

+ + +
+

r ) h r (
2
r MG 2 r MG 2 r r
) r r (
) (# T (7.37)

2 / 1
1
h r
r r
r r
h r
sin a

,
_

+
+
(7.38)

+
+

r ) h r (
h ) r r (
m
2
(7.39)
8. Mercurys Perihelion Shift
Mercurys perihelion shift (per revolution) is given by (ref. [1])
a 2 | ) r ( ) r ( | 2 A

(8.1)
From (7.36) we get, since
2
a
a (from (7.38))
) m ( K ) (# T
sin m 1
d
) (# T ) r ( ) r (
2 / a
0
2 2

+
(8.2)
We thus find
a 2 ) m ( K ) (# T 2 a 2 | ) r ( ) r ( | 2 A

(8.3)
9. Trajectory of a Light Signal.
9.1 Trajectory of Light Signal using Weirstrass Elliptic Function.
For a light signal we have 0 E . We get from equation (4.6)

) r ( A
r
J
r
d
dr
2
2
4
2

,
_

(9.1)
where

) r ( B
r
J
0
2
0
2
,
0
0
r
MG 2
1 ) r ( B ,
1
r
MG 2
1 ) r ( A

,
_

(9.2)
Using relations (9.2), we get from (9.1)

2 4
2
0
0
2
r
r
MG 2
1 r
r
) r ( B
d
dr

,
_

,
_

27 Page
which is equivalent to
r MG 2 r r
r
MG 2 r
d
dr
2 4
3
0
0
2
+

,
_

(9.3)
The change of variables

r
1
u ,
d
du
u
1
d
dr
2
(9.4)
transforms equation (9.3) into

u
MG 2
u
1
u
1
r
MG 2 r
d
du
u
1
2 4 3
0
0
2
2
+

,
_

which is equivalent to

3
0
0
2 3
2
r
MG 2 r
u u MG 2
d
du

+

,
_

(9.5)
Using the linear transformation

MG 6
1
U
MG
2
u + ,
d
dU
MG
2
d
du
(9.6)
equation (9.5) becomes

,
_

,
_

3
2
MG 6
1
U
MG
2
MG 2
d
dU
MG
2

3
0
0
2
r
MG 2 r
MG 6
1
U
MG
2

+

,
_

+
which is equivalent to

3 2
3
2
g U g U 4
d
dU

,
_

(9.7)
where

12
1
g
2
(9.8)
28 Page

3
0
0
2 2
3
r 4
) MG 2 r ( G M
216
1
g

(9.9)
The solution of (9.7) is given by
) g , g ; c ( U
3 2
+ (9.10)
where is the Weirstrass elliptic function and c is a constant. We thus arrive at
the following solution of equation (9.1):

MG 6
1
) g , g ; c (
MG
2
r
1
3 2
+ + (9.11)
9.2 Trajectory of Light Signal using Jacobis Elliptic Function.
First Case: Solution of the equation (9.5)
Equation (9.5) is equivalent to
) e u )( e u )( e u ( MG 2
d
du
3 2 1
2

,
_

(9.12)
where
1
e ,
2
e ,
3
e are the three roots of the equation
0
r
MG 2 r
u u MG 2
3
0
0
2 3

+
given by

0
0 0 0
1
r MG 4
) MG 6 r )( MG 2 r ( MG 2 r
e
+ +

0
2
r
1
e

0
0 0 0
3
r MG 4
) MG 6 r )( MG 2 r ( MG 2 r
e
+

with the ordering

3 2 1
e e e > >
Using the substitution
29 Page

3 2
3 2
e e
e u
z

3 1
3 2 2
e e
e e
h

( 1 h 0
2
< < )

3 1
2
e e
equation (9.12) becomes
) z h 1 )( z 1 (
dx
dz
2 2 2
2

,
_

(9.13)
where
t x ,
2
) e e ( MG

2
MG
t
3 1


The general solution of the equation (9.13) is given by
) h ; o x ( sn z + , o constant
and so
) h ; o t ( sn ) e e ( e u
2
3 2 3
+ +
The equation of the orbit is thus written as

) h ; o t ( sn ) e e ( e
1
r
2
3 2 3
+ +
(9.14)
We may calculate the constant o in the previous formula by imposing some initial
conditions.
Second Case: Solution of the equation (9.7)
Equation (9.7) is equivalent to
) e
~
U )( e
~
U )( e
~
U ( MG 4
d
dU
3 2 1
2

,
_

(9.15)
where
1
e
~
,
2
e
~
,
3
e
~
are the three roots of the equation 0 g U g U 4
3 2
3
, given
by
30 Page

0
0 0 0
1
r 24
) MG 6 r )( MG 2 r ( 3 MG 6 r
e
~
+ +

0
0
2
r 12
r MG 6
e
~

0
0 0 0
3
r 24
) MG 6 r )( MG 2 r ( 3 MG 6 r
e
~
+

with the ordering

3 2 1
e
~
e
~
e
~
> > under the further assumption MG 6 r
0
> .
Under the substitution

3 2
3 2
e
~
e
~
e
~
U
z

3 1
3 2 2
e
~
e
~
e
~
e
~
g

( 1 g 0
2
< < )

3 1
2
e
~
e
~
u
equation (9.15) becomes
) z g 1 )( z 1 (
dx
dz
2 2 2
2

,
_

(9.16)
where
u x ,
3 1
e
~
e
~
u
The general solution of the equation (9.16) is given by
) g ; o x ( sn z + , o constant
and so
) g ; o u ( sn ) e
~
e
~
( e
~
U
2
3 2 3
+ +
Using (9.6), we arrive at the following equation of the orbit:

MG 6
1
)] g ; o u ( sn ) e
~
e
~
( e
~
[
MG
2
r
1
2
3 2 3
+ + + (9.17)
31 Page
10. Period of Revolution
We shall now calculate the period of revolution of a test body describing a closed
orbit around a heavy object of mass M producing the gravitational field. We start
with equation (A.14) of Appendix A, written in equivalent form
) x ( R
dr
dt
(10.1)
where

2 / 1 3 2 3
] x ) x ( B E x ) x ( B J x [
x x
MG 2 x
x
) x ( R

(10.2)
The period of revolution is been expressed as

p
r
r
r
dx ) x ( R dx ) x ( R 2 T (10.3)
where
) a 2 ( r p (10.4)
We have, according to (7.32):


3 2 3
x ) x ( B E x ) x ( B J x
)} h x )( r x )( r x ( {
) MG 2 r )( MG 2 r )( r r (
) r MG 2 r MG 2 r r ( MG 2

+

+
+ +
+ +
(10.5)
Because of (10.5), we get from (10.2) that

) h x )( r x )( r x ( ) MG 2 x (
x x
) (# P ) x ( R
2

+
(10.6)
where

) r MG 2 r MG 2 r r ( MG 2
) MG 2 r )( MG 2 r )( r r (
) (# P
+ +
+ +

+
(10.7)
We observe that

) x ( H ) MG 2 x (
x
) (# P ) x ( R
3

(10.8)
32 Page
where
) h x )( r x )( r x ( x ) x ( H
+
(10.9)
According to Appendix F, under the substitution

h x
r x
r r
h r
sin
2

+
+
(10.10)

sin ) r r ( ) h r (
sin ) r r ( h ) h r ( r
x
2
2
+ +
+ +


(10.11)
we have

sin m 1
d
r ) h r (
2
) x ( H
dx
2 2

+
(10.12)
where

+
+

r ) h r (
h ) r r (
m
2
(10.13)
In Appendix G, under the same substitution (10.11) and introducing further the
notation

h r
r r
k
2

+
+
(10.14)
and

) MG 2 r )( h r (
) MG 2 h )( r r (
q

+
+
( 1 q 0 < < ) (10.15)
we find that

dx
) x ( H ) MG 2 x (
x
3
d
sin m 1 ) sin q 1 ( ) sin k 1 (
) sin m 1 (
h r
r
MG 2 r
) r ( 2
2 2 2 2 2 2
3 2 2 2

(10.16)
and finally
d ) ( S ) (# Q dx ) x ( R (10.17)
33 Page
where
) r MG 2 r MG 2 r r )( h r )( MG 2 r ( MG 2
r ) MG 2 r )( r r (
) r ( 2 ) (# Q
2
+ + +
+ +

+
(10.18)
d
sin m 1 ) sin q 1 ( ) sin k 1 (
) sin m 1 (
d ) ( S
2 2 2 2 2 2
3 2 2

(10.19)
We now find the new limits of the integrals. Since

h x
r x
r r
h r
sin
2

+
+
and
sin ) r r ( ) h r (
sin ) r r ( h ) h r ( r
x
2
2
+ +
+ +

we find

+
r x :
2
a
1 sin
2

r x : 0 0 sin
2

and
p x : a
h p
r p
r r
h r
sin
h p
r p
r r
h r
sin
2 / 1
1 2

,
_

+
+
+
+

r x : 0
Using (10.3) and (10.17) we get the following formula for the period:

1
1
]
1

a
0
2 / a
0
d ) ( S d ) ( S 2 ) (# Q T (10.20)
In order to calculate the period, we have to calculate the two integrals appearing in
(10.20). For this purpose, we first have to use the identity (Appendix H)

) sin q 1 ( ) sin k 1 (
) sin m 1 (
2 2 2 2
3 2 2
+

'

+

sin k 1
1
) k q ( k
k q k 2 m k 2 q m
) k m (
q k
m
2 2
2 2 4
4 2 2 2 2
2 2 2
4
6
34 Page

) sin q 1 ( ) k q ( q
) q m (
sin k 1
1
k ) k q ( k
) k m (
2 2 2
3 2
2 2 2 2 2
3 2 2

,
_

+ (10.21)
Using the previous partial fraction decomposition we find, using equation (10.20)
that the period is given by

1
1
]
1

sin m 1
d
2 ) (# Q T
2 2
a
0
2 / a
0
1
+

1
1
]
1

sin m 1 ) sin k 1 (
d
2 ) (# Q
2 2 2 2
a
0
2 / a
0
2


1
1
]
1

+

sin m 1 ) sin k 1 (
d
2
k
) (# Q
2 2 2 2
a
0
2 / a
0
2
3

sin m 1 ) sin q 1 (
d
2 ) (# Q
2 2 2
a
0
2 / a
0
4

1
1
]
1

(10.22)
where
) (# Q
q k
m
) (# Q
4
6
1

) (# Q
) k q ( k
k q k 2 m k 2 q m
) k m ( ) (# Q
2 2 4
4 2 2 2 2
2 2 2
2

+

) (# Q
) k q ( k
) k m (
) (# Q
2 2
3 2 2
3

) (# Q
) k q ( q
) q m (
) (# Q
2 2
3 2
4

We make further use of the substitutions

2 / a
0
2 2
sin m 1
d
) m ( K (10.23)
35 Page

a
0
2 2
sin m 1
d
) m ; a ( F (10.24)

2 / a
0
2 2 2 2
2
sin m 1 ) sin k 1 (
d
) m ; k ,
2
a
( H (10.25)

a
0
2 2 2 2
2
sin m 1 ) sin k 1 (
d
) m ; k , a ( H (10.26)

2 / a
0
2 2 2
sin m 1 ) sin q 1 (
d
) m ; q ,
2
a
( H (10.27)

a
0
2 2 2
sin m 1 ) sin q 1 (
d
) m ; q , a ( H (10.28)
to convert equation (10.22) into a more compact form.
)} m ; a ( F ) m ( K 2 { ) (# Q T
1
+ )} m ; k , a ( H ) m ; k ,
2
a
( H 2 { ) (# Q
2 2
2

+ )} m ; k , a ( H ) m ; k ,
2
a
( H 2 {
k
) (# Q
2 2
2
3
)} m ; q , a ( H ) m ; q ,
2
a
( H 2 { ) (# Q
4
(10.29)
We have now to find the quantity ) a 2 ( r p . In section (7.2) we have found
that the equation of the orbit is given by

) k ; ( sn ) e e ( e
1
r
2
3 2 3
+

where

h
1
) r r ( MG 2
) r r ( MG 2 r r
e
1

+

+
+ +
,

r
1
e
2
,
+

r
1
e
3
(
3 2 1
e e e > > )
are the three roots of the equation
36 Page
0 u MG 2 u u
b
aMG 2
b
a c
3 2
+ +

and

2
) e e ( MG

3 1

In obtaining the equation of the orbit we have considered the initial condition
0 for
+
r r
We thus find that

) k ; a 2 ( sn ) e e ( e
1
) a 2 ( r p
2
3 2 3
+
(10.30)
11. Time Delay of Radio Signals in the Suns
Gravitational Field.
The time taken for the light to reach a star P whose distance from the Sun is
P
r
and received back to Earth by the emitting device, is given by
)} r , r ( t ) r , r ( t { 2 T
0 P 0
+

(11.1)
where

r is Earths distance from the Sun and

0
r is the closest approach of the
orbit of light from the Sun (Ref. [4]).
We are going now to calculate the quantity ) r , r ( t
0
which represents the time
taken by the light to go from a point whose distance is r to a point whose distance
is
0
r from a massive point of mass M, which produces the gravitational field.
Since for radio signals 0 E , we get from equation (4.10)

'

,
_

2
2 2
2
r
J
) r ( B
1
) r ( A
) r ( B
dt
dr
which is equivalent to

)] r ( B J r [ ) r ( B
r
dr
dt
2 2 2
2
2

,
_

37 Page
from which we get

2 / 1 2 2
)] r ( B J r [ ) r ( B
r
dr
dt

(11.2)
Since

) r ( B
r
J
0
2
0
2
(11.3)
we finally get from (11.2):

2 / 1 2
0 0
2
0
)] r ( B r ) r ( B r [ ) r ( B
) r ( B r
dr
dt

(11.4)
Separating the variables in the previous equation and integrating, we get

r
r
2 / 1 2
0 0
2
0
0
0
dx
)] x ( B r ) r ( B x [ ) x ( B
) r ( B x
) r , r ( t (11.5)
Since, as follows from (6.6)

2 / 1 2
0 0
2
)] x ( B r ) r ( B x [ x
) r x )( r x )( r x ( x
r
MG 2 r
2 1 0
0
0

(11.6)
and

MG 2 x
x
r
MG 2 r
x
MG 2
1
r
MG 2
1
) x ( B
) r ( B
0
0 0 0

(11.7)
we get from (11.5)

r
r
2 1 0
3
0
0
dx
) r x )( r x )( r x ( x ) MG 2 x (
x
) r , r ( t (11.8)
In Appendix I we find, under the substitution
38 Page

sin ) r r ( ) r r (
sin ) r r ( r ) r r ( r
x
2
1 0 1 2
2
1 0 2 1 2 0

that

) r x )( r x )( r x ( x ) MG 2 x (
x
2 1 0
3
d
sin k 1 ) sin p 1 ( ) sin m 1 (
) sin k 1 (
) (# U
2 2 2 2 2 2
3 2 2
+

(11.9)
where

) r r ( r
) r r ( r
k
1 2 0
1 0 2 2

(11.10)

1 2
1 0 2
r r
r r
m

(11.11)

1 2
1 0
0
2
r r
r r
MG 2 r
MG 2 r
p

(11.12)
and

4 / 1
0
0
0
2
0
1 2
0
0
2
0
MG 6 r
MG 2 r
MG 2 r
r 2
r r
r
MG 2 r
r 2
) (# U

,
_

(11.13)
We calculate now the new limits of integration from (I.12) (Appendix I)
0 0 sin r x
2
0

a
r r
r r
r r
r r
sin
r r
r r
r r
r r
sin r x
2 / 1
2
0
1 0
1 2 1
2
0
1 0
1 2 2

,
_

(11.14)
The integral in (11.8) is converted into

r
r
2 1 0
3
0
0
dx
) r x )( r x )( r x ( x ) MG 2 x (
x
) r , r ( t
39 Page

a
0
2 2 2 2 2 2
3 2 2
d
sin k 1 ) sin p 1 ( ) sin m 1 (
) sin k 1 (
) (# U (11.15)
We can convert the integral in (11.15) into a combination of Elliptic
Integrals. For this purpose we use the identity (J.4), Appendix J.

) sin p 1 ( ) sin m 1 (
) sin k 1 (
2 2 2 2
3 2 2

+
+
sin m 1
1
) p m ( m
) p m 2 k p m k 2 m ( ) m k (
p m
k
2 2 2 2 4
2 2 2 2 4 2 2 2
4
6
sin p 1
1
) p m ( p
) p k (
sin m 1
1
m ) p m ( m
) m k (
2 2 2
3 2
2 2 2 2 2
3 2 2

,
_

(11.16)
We thus find the following expression

a
0
2 2
1 0
sin k 1
d
) (# U ) r , r ( t
+

+

a
0
2 2 2 2
2
sin k 1 ) sin m 1 (
d
) (# U
+

+

a
0
2 2 2 2
2
3
sin k 1 ) sin m 1 (
d
m
) (# U

+
+
a
0
2 2 2
4
sin k 1 ) sin p 1 (
d
) (# U (11.17)
where
) (# U
p m
k
) (# U
4
6
1

) (# U
) p m ( m
p m 2 pk m k 2 m
) m k ( ) (# U
2 2 4
2 2 2 2 4
2 2 2
2

'

+

40 Page
) (# U
) p m (
) k m (
m
1
) (# U
2
3 2 2
2
3

) (# U
) p m ( p
) k p (
) (# U
2 2
3 2

Introducing the notation

a
0
2 2
sin k 1
d
) k ; a ( F

a
0
2 2 2 2
2
sin k 1 ) sin m 1 (
d
) k ; m , a ( H

a
0
2 2 2
sin k 1 ) sin p 1 (
d
) k ; p , a ( H
we obtain the expression
+ + ) k ; m , a ( H ) (# U ) k ; a ( F ) (# U ) r , r ( t
2
2 1 0
) k ; p , a ( H ) (# U ) k ; m , a ( H
m
) (# U
4
2
2
3
+

+ (11.18)
Using (11.18), we obtain the following formula for the time delay:
+

)} r , r ( t ) r , r ( t { 2 T
0 P 0
+ + )} k ; q ( F ) k ; ( F ){ (# U 2
1
+ + + )} k ; m , q ( H ) k ; m , ( H { ) (# U 2
2 2
2
+ +

+ )} k ; m , q ( H ) k ; m , ( H {
m
) (# U 2
2 2
2
3
)} k ; p , q ( H ) k ; p , ( H { ) (# U 2
4
+ + (11.19)
where

2 / 1
2
0
1 0
1 2 1
r r
r r
r r
r r
sin

,
_

(11.20)
41 Page
and

2 / 1
2 P
0 P
1 0
1 2 1
r r
r r
r r
r r
sin q

,
_

(11.21)
Appendix A. Simplification of the Equations of Motion.
Step 1. We divide equation (3.9) by
dp
dt
and we find
0
dp
dr
) r ( B
) r ( B
dp
t d
dp
dt
1
2
2

+
which is successively equivalent to

,
_

+ + 0 B ln
dp
dt
ln
dp
d
0 B ln
dp
d
dp
dt
ln
dp
d

,
_

,
_

1
C B
dp
dt
ln 0 B
dp
dt
ln
dp
d
C B
dp
dt

We choose to normalize p so that the solution of the last equation should look
as

) r ( B
1
dp
dt
(A.1)
Step 2. We divide equation (3.8) by
dp
d
and we find
0
dp
dr
r
2
dp
d
dp
d
1
2
2
+
which is successively equivalent to

,
_

+ + 0 r ln
dp
d
ln
dp
d
0 r ln
dp
d
dp
d
ln
dp
d
2 2
42 Page

,
_

,
_

2
2 2
C r
dp
d
ln 0 r
dp
d
ln
dp
d
J
dp
d
r
2
(A.2)
where J is a constant having dimensions of angular momentum per unit mass.
Step 3. We substitute
dp
dt
and
dp
d
as given by equations (A.1) and (A.2) into
equation (3.7) and we find
0
) r ( B
1
) r ( A 2
) r ( B
r
J
) r ( A
r
dp
dr
) r ( A 2
) r ( A
dp
r d
2 2
2
2
2
2

,
_

,
_

,
_

+
and multiplying by
dp
dr
) r ( A 2 we find
0
dp
dr
) r ( B
) r ( B
dp
dr
r
J
2
dp
dr
) r ( A
dp
r d
dp
dr
) r ( A 2
2 3
2
3
2
2

,
_

+
which is successively equivalent to

,
_

,
_

,
_

,
_

0
) r ( B
1
dp
d
r
J
dp
d
dp
dr
) r ( A
dp
d
2
2
2

,
_

,
_

0
) r ( B
1
r
J
dp
dr
) r ( A
dp
d
2
2
2
E
) r ( B
1
r
J
dp
dr
) r ( A
2
2
2
+

,
_

(A.3)
where E is a constant.
Step 4. Relation between proper time and affine parameter p.
We start from

2 2 2 2 2 2 2 2
d 0 sin r 0 d r dr ) r ( A dt ) r ( B t d (2.1)
43 Page
and use the fact that
2
a
0 so as equation (2.1) takes the form

2 2 2 2 2
d r dr ) r ( A dt ) r ( B t d
which can be written as

2
2
2 2
2
2
dp
d
r
dp
dr
) r ( A
dp
dt
) r ( B
dp
t d

,
_

,
_

,
_

(A.4)
We substitute in the above equation the derivatives
dp
dt
and
dp
d
as given by
equations (A.1) and (A.2) respectively. We obtain the equation

2
2
2
2
2
r
J
dp
dr
) r ( A
) r ( B
1
dp
t d

,
_

(A.5)
We easily recognize that the rhs of the above equation is E, because of equation
(A.3). This means that

2 2
dp E t d (A.6)
where
0 E > for particles
0 E for photons
Step 5. Equation of the orbit ) ( r r .
The equation of the orbit is obtained by eliminating dp from the equations (A.2)
and (A.3). In fact using d
J
r
dp
2
, equation (A.3) takes one of the following
equivalent forms

2 2 2
2
4
J
E
) r ( B J
1
r
1
d
dr
r
) r ( A
+

,
_

(A.7)
or
44 Page

,
_

,
_

2 2 2
4
2
r
1
J
E
) r ( B J
1
) r ( A
r
d
dr
(A.8)
or
1
r
J
E
) r ( A
1
d
dr
r
J
2
2
2
2

,
_

+ +

,
_

(A.9)
Step 6. The value of J.
6.1 Suppose the orbit is an unbounded curve. At the closest distance from the
object which produces the gravitational field, we have
0
r r and so 0
d
dr
.
Therefore equation (A.7) gives

2
0
2 2
0
J
E
) r ( B J
1
r
1

from which, since
2
V 1 E (A.10)
we get

,
_

+
2
0
2
0
2
V 1
) r ( B
1
r J (A.11)
6.2 When the path is a closed curve, the value of J is not longer given by
equation (A.11), because in this case there are two points for which 0
d
dr
.
Let us call
t
r these points. Using equation (A.7) we get
0 E
) r ( B
1
r
J
2
2
+
t
t
(A.12)
In other words we have the system
0 E
) r ( B
1
r
J
2
2
+
+
+
(A.13a)
45 Page
0 E
) r ( B
1
r
J
2
2
+

(A.13b)
which is solved in Appendix D.
Step 7.
Since dt ) r ( B dp (equation (A.1)), equation (A.3) gives us
E
) r ( B
1
r
J
) r ( B
1
dt
dr
) r ( A
2
2
2
+

,
_

(A.14)
We may simplify further equation (A.14). We find

)] r ( B r E ) r ( B J r [ ) r ( B
) r ( A r
dr
dt
2 2 2
2
2

,
_

from which we obtain

2 / 1 3 2 3
)] r ( B r E ) r ( B r J r [
r r
) r ( B
) r ( A
dr
dt

(A.15)
We put

2 / 1 3 2 3
)] x ( B x E ) x ( B x J x [
x x
) x ( B
) x ( A
) x ( R

(A.16)
We also have

2
2
2 2
) MG 2 x (
x
x
MG 2
1
1
) x ( B
) x ( B ) x ( A
) x ( B
) x ( A

,
_

from which we get

MG 2 x
x
) x ( B
) x ( A

(A.17)
Therefore the expression for ) x ( R is given by

2 / 1 3 2 3
)] x ( B x E ) x ( B x J x [
x x
MG 2 x
x
) x ( R

(A.18)
46 Page
Equation (A.15) is equivalent to
) r ( R
dr
dt
(A.19)
with ) x ( R given by (A.18).
Appendix B.
In this Appendix we prove identity (6.6). We have
)] x ( B r ) r ( B x [ x )] x ( B r ) r ( B x [ x
2
0 0
2 2 2 / 1 2
0 0
2

1
]
1

,
_

,
_

x
MG 2
1 x r
r
MG 2
1 x x
2
0
0
3

1
]
1

,
_

2
0
2
0
3
0
0
r MG 2 x r x
r
MG 2 r
x

1
1
]
1

MG 2 r
r MG 2
x
MG 2 r
r
x x
r
MG 2 r
0
3
0
0
3
0 3
0
0
(B.1)
One of the roots of the cubic polynomial which appears in the bracket under the
square root of (B.1) is
0
r and so

,
_

MG 2 r
r MG 2
x r x ) r x (
MG 2 r
r MG 2
x
MG 2 r
r
x
0
2
0
0
2
0
0
3
0
0
3
0 3

The discriminant of the quadratic trinomial inside the parenthesis is

MG 2 r
) MG 6 r ( r
MG 2 r
r MG 8
r D
0
0
2
0
0
2
0 2
0

+
Suppose that MG 2 r
0
> and the real roots are given by

,
_

+
+
MG 2 r
MG 6 r
1 r
2
1
r
0
0
0 1
and

,
_

+

MG 2 r
MG 6 r
1 r
2
1
r
0
0
0 2
Therefore we have
47 Page

2 / 1 2
0 0
2
)] x ( B r ) r ( B x [ x
) r x ( ) r x ( ) r x ( x
r
MG 2 r
2 1 0
0
0

(B.2)
We have to mention that
2 0
r r > if MG 3 r
0
> (provided that MG 2 r
0
> ).
So instead of MG 2 r
0
> we may impose the stronger condition MG 3 r
0
> and
under this assumption we have

1 2 0
r r r > >
and under the assumption
3 3 2
0
G M 32 r > we have the ordering

1 2 0
r r MG 2 MG 3 r > > > > (B.3)
Appendix C. In this Appendix we shall express the integral in (6.10) in terms
of elliptic integrals. We put

0 1
r a ,
2 2
r a , 0 a
3
and
1 4
r a (C.1)
Since
2 0
r r > and
1
r 0 > , we have

2 1
a a > and
4 3
a a > (C.2)
Under these assumptions, the substitution (Ref. [3])

sin a a
sin a a a a
x
2
41 42
2
41 2 42 1

(C.3)
implies the relation

sin k 1
d

) x ( G
dx
2 2

(C.4)
where

4 2
3 2
3 1
4 1 2
a a
a a
a a
a a
k

(C.5)
and
48 Page

42 31
a a
2
(C.6)
Since

1 2 4 2 42
r r a a a ,
1 0 4 1 41
r r a a a (C.7)
and

0 3 1 31
r a a a (C.8)
we find

) r r ( r
2

1 2 0

(C.9)
and

1 2
1 0
0
2 2
r r
r r
r
r
k

(C.10)

sin ) r r ( ) r r (
sin ) r r ( r ) r r ( r
x
2
1 0 1 2
2
1 0 2 1 2 0


(C.11)
We have thus found that

sin k 1
d
) r r ( r
2
) x ( G
dx
2 2
1 2 0

We also have the following relation

2
0
1 0
1 2
2
1
41
42 2
r x
r x
r r
r r
a x
a x
a
a
sin

(C.12)
From the previous equation we can determine the new limits of the integral:
0 0 sin r x
2
0

a
r r
r r
sin
r r
r r
sin x
2 / 1
1 0
1 2
1
1 0
1 2
2

,
_

(C.13)
In this way we have found that
49 Page

a
0
2 2
1 2 0
r
sin k 1
d
) r r ( r
2
) x ( G
dx
0
(C.14)
Appendix D. In this Appendix we solve the system of equations (7.21).
Subtracting the two equations (7.21) we get

) r ( B
1
) r ( B
1
r
1
r
1
J
2 2
2
+
+

,
_

from which we obtain

) r r )( r r (
) r r (
) r ( B ) r ( B
) r ( B ) r ( B
J
2
2
+ +
+
+
+
+

(D.1)
Adding the two equations (7.21) we get

,
_

+ +
+
+
2 2
2
r
1
r
1
J
) r ( B
1
) r ( B
1
E 2
from which, using the expression (D.1) for
2
J , we find

) r r )( r ( B ) r ( B
r ) r ( B r ) r ( B
E
2 2
2 2
+ +
+ +

(D.2)
We have now to simplify further the two expressions we have derived above for
2
J and E. Using the expression (2.2) for ) r ( B we obtain

) r r )( MG 2 r )( MG 2 r (
) r r ( MG 2
J
2
2
+ +
+
+
(D.3)
and

) r r )( MG 2 r )( MG 2 r (
) r r ( MG 2 ) MG 2 r ( ) r ( ) MG 2 r ( ) r (
E
2 2
+ +
+ + +
+
+
(D.4)
Appendix E. In this Appendix we prove (7.32).
Using (7.30) and (7.31), we have
x ) x ( B J x ) x ( B E x
2 3 3
50 Page

,
_

,
_

x
x
MG 2
1
c
b
x
x
MG 2
1
c
a
x
3 3

c
bMG 2
x
c
b
x
c
MG a 2
x
c
a
1
2 3
+ +

,
_

(E.1)
The roots of the cubic polynomial which appears in (E.1) are

+
r ,

r and
+ +
+

r MG 2 r MG 2 ) r r (
) r r ( MG 2
: h (E.2)
We also make the assumption
MG 2 h r r > > >
+
and MG 6 r >

Since
) r MG 2 r MG 2 r r ( MG 2 a c
+ +

using (E.1) and (E.2) we obtain
x ) x ( B J x ) x ( B E x
2 3 3
)} h x )( r x )( r x ( {
) MG 2 r ( ) MG 2 r ( ) r r (
) r MG 2 r MG 2 r r ( MG 2

+

+
+ +
+ +
(E.3)
Appendix F.
We shall express the integral in (7.35) in terms of an elliptic integral.
We put

+
r a
1
,

r a
2
, h a
3
and 0 a
4
(F.1)
Since
+
> r r and 0 h > , we have

2 1
a a > and
4 3
a a > (F.2)
Under these assumptions, the substitution (Ref. [3])

sin a a
sin a a a a
x
2
21 31
2
21 3 31 2

(F.3)
implies the relation
51 Page

sin m 1
d

) x ( H
dx
2 2

(F.4)
where
) h x ( ) r x ( ) r x ( x ) x ( H
+
(F.5)

4 2
1 2
1 3
4 3 2
a a
a a
a a
a a
m

(F.6)

42 31
a a
2
(F.7)
We also have

3
2
21
31 2
a x
a x
a
a
sin

(F.8)
Since
h r a a a
3 1 31

+

+
r r a a a
2 1 21
(F.9)

r a a a
4 2 42

we find

+

r ) h r (
2
(F.10)

h x
r x
r r
h r
sin
2

+
+
(F.11)

+
+

r ) h r (
h ) r r (
m
2
(F.12)

sin ) r r ( ) h r (
sin ) r r ( h r ) h r (
x
2
2
+ +
+ +


(F.13)
We thus arrive at

sin m 1
d
r ) h r (
2
) h x ( ) r x ( ) r x ( x
dx
2 2

+ +
(F.14)
52 Page
For the limits we find
0 0 sin r x
2


a
h r
r r
r r
h r
sin
h r
r r
r r
h r
sin r x
2 / 1
1 2

,
_

+
+
+
+
We thus have found

+ +

a
0
2 2
r
r
sin m 1
d
r ) h r (
2
) h x ( ) r x ( ) r x ( x
dx
(F.15)
Appendix G. Introducing the notation

h r
r r
k
2

+
+
(G.1)
we get from (10.11) that

sin k 1
r ) sin m 1 (
x
2 2
2 2

(G.2)
and then

sin k 1
sin ) k MG 2 r m ( ) MG 2 r (
MG 2 x
2 2
2 2 2

(G.3)
We have further

+
+

+
+

h r
r r
MG 2 r
r ) h r (
h ) r r (
k MG 2 r m
2 2

h r
) MG 2 h ( ) r r (

+
+
(G.4)
and then we get from (G.3) that

sin k 1
sin q 1
) MG 2 r ( MG 2 x
2 2
2

(G.5)
where
53 Page

) MG 2 r )( h r (
) MG 2 h )( r r (
q

+
+
( 1 q 0 < < ) (G.6)
Using (10.12), (G.2) and (G.5), we obtain

dx
) x ( H ) MG 2 x (
x
3

sin m 1 ) sin q 1 ( ) sin k 1 (
d ) sin m 1 (
h r
r
MG 2 r
) r ( 2
2 2 2 2 2 2
3 2 2 2

(G.7)
Using (10.8) and (10.16), we have
dx
) x ( H ) MG 2 x (
x
) (# P dx ) x ( R
3

and using (G.7) we arrive at
d ) ( S ) (# Q dx ) x ( R (G.8)
where

+ + +
+ +

r MG 2 r MG 2 r r )( h r ( ) MG 2 r ( MG 2
r ) MG 2 r )( r r (
) r ( 2 ) (# Q
2
(G.9)
and

sin m 1 ) sin q 1 ( ) sin k 1 (
d ) sin m 1 (
d ) ( S
2 2 2 2 2 2
3 2 2

(G.10)
Appendix H. In this Appendix we prove the identity (10.21).
For this purpose, we first have to make the partial fraction expansion

) x q 1 ( ) x k 1 (
) x m 1 (
2 2
3 2
+

+

x k 1
1
) k q ( k
) q k m k 3 q m 2 ( ) k m (
q k
m
2 2 2 4
2 2 2 2 2 2 2
4
6

x q 1
1
) k q ( q
) q m (
) x k 1 (
1
) k q ( k
) k m (
2 2
3 2
2 2 2 4
3 2 2

+ (H.1)
54 Page
Using the identity

,
_

x k 1
1
k
k
x k 1
1
) x k 1 (
1
2 2
2
2 2 2
(H.2)
relation (H.1) becomes

) x q 1 ( ) x k 1 (
) x m 1 (
2 2
3 2
+

+
+
x k 1
1
) k q ( k
) q k 2 q m m k 2 k ( ) k m (
q k
m
2 2 2 4
2 2 2 2 4 2 2 2
4
6

x q 1
1
) k q ( q
) q m (
x k 1
1
k ) k q ( k
) k m (
2 2
3 2
2 2 2 2
3 2 2

,
_

+ (H.3)
Substituting sin x
2
in (H.3) we arrive at the identity (10.21).
Appendix I.
We put

0 1
r a ,
2 2
r a , 0 a
3
and
1 4
r a (I.1)
Since
2 0
r r > and
1
r 0 > , we have

2 1
a a > and
4 3
a a > (I.2)
Under these assumptions, the substitution (Ref. [3])

sin a a
sin a a a a
x
2
41 42
2
41 2 42 1

(I.3)
implies the relation

sin k 1
d

) x ( G
dx
2 2

(I.4)
where

4 2
3 2
3 1
4 1 2
a a
a a
a a
a a
k

(I.5)
and
55 Page

42 31
a a
2
(I.6)
Since

1 2 4 2 42
r r a a a ,
1 0 4 1 41
r r a a a (I.7)
and

0 3 1 31
r a a a (I.8)
we find

) r r ( r
2

1 2 0

(I.9)

1 2
1 0
0
2 2
r r
r r
r
r
k

(I.10)

sin ) r r ( ) r r (
sin ) r r ( r ) r r ( r
x
2
1 0 1 2
2
1 0 2 1 2 0


(I.11)
We also have the following relation

2
0
1 0
1 2
2
1
41
42 2
r x
r x
r r
r r
a x
a x
a
a
sin

(I.12)
Relation (I.4) takes on the final form

sin k 1
d
) r r ( r
2
) x ( G
dx
2 2
1 2 0

(I.13)
We make the substitutions
sin ) r r ( r ) r r ( r ) ( G
2
1 0 2 1 2 0
) sin k 1 )( r r ( r
2 2
1 2 0
(I.14)
and
sin ) r r ( ) r r ( ) ( F
2
1 0 1 2
) sin m 1 )( r r (
2 2
1 2
(I.15)
where
56 Page

) r r ( r
) r r ( r
k
1 2 0
1 0 2 2

(I.16)

1 2
1 0 2
r r
r r
m

(I.17)
Because of the previous substitutions (I.14) and (I.15), we get, using (I.14)-(I.17)
that (I.11) becomes

sin m 1
) sin k 1 ( r
x
2 2
2 2
0

(I.18)
Under the substitution (I.18), we find

1
1
]
1

MG 2
sin m 1
) sin k 1 ( r
) sin m 1 (
) sin k 1 ( r
MG 2 x
x
2 2
2 2
0 3 2 2
3 2 2 3
0
3
)] sin m 1 ( MG 2 ) sin k 1 ( r [ ) sin m 1 (
) sin k 1 ( r
2 2 2 2
0
2 2 2
3 2 2 3
0

(I.19)
We have further
) sin m 1 ( MG 2 ) sin k 1 ( r
2 2 2 2
0
sin ) m MG 2 k r ( ) MG 2 r (
2 2 2
0 0

'

sin
MG 2 r
m MG 2 k r
1 ) MG 2 r (
2
0
2 2
0
0
(I.20)
and since

1 2
1 0
1 2
1 0 2 2 2
0
r r
r r
MG 2
r r
) r r ( r
m MG 2 k r
) MG 2 r (
r r
r r
2
1 2
1 0

we get from (I.20)

57 Page
) sin m 1 ( MG 2 ) sin k 1 ( r
2 2 2 2
0
) sin p 1 )( MG 2 r (
2
0
(I.21)
where

1 2
1 0
0
2
r r
r r
MG 2 r
MG 2 r
p

(I.22)
Using (I.21), we get from (I.19):

) sin p 1 ( ) sin m 1 (
) sin k 1 (
MG 2 r
r
MG 2 x
x
2 3 2 2
3 2 2
0
3
0
3

Finally, collecting everything together, we have

) r x )( r x )( r x ( x ) MG 2 x (
x
2 1 0
3
d
sin k 1 ) sin p 1 ( ) sin m 1 (
) sin k 1 (
) (# U
2 2 2 2 2 2
3 2 2

(I.23)
where

1 2
0
0
2
0
r r
r
MG 2 r
r 2
) (# U

(I.24)
Appendix J. In this Appendix we prove identity (11.16).
For this purpose we use the partial fraction decomposition

) x p 1 ( ) x m 1 (
) x k 1 (
2 2
3 2

+
x m 1
1
) p m ( m
) p m k p 2 m k 3 ( ) m k (
p m
k
2 2 2 4
2 2 2 2 2 2 2
4
6
x p 1
1
) p m ( p
) p k (
) x m 1 (
1
) p m ( m
) m k (
2 2
3 2
2 2 2 4
3 2 2

(J.1)
Using the identity
58 Page

,
_

x m 1
1
m
m
x m 1
1
) x m 1 (
1
2 2
2
2 2 2
(J.2)
relation (J.1) takes the form

) x p 1 ( ) x m 1 (
) x k 1 (
2 2
3 2

+
+
x m 1
1
) p m ( m
) p m 2 pk m k 2 m ( ) m k (
p m
k
2 2 2 4
2 2 2 2 4 2 2 2
4
6
x p 1
1
) p m ( p
) p k (
x m 1
1
m ) p m ( m
) m k (
2 2
3 2
2 2 2 2
3 2 2

,
_

(J.3)
After the substitution sin x
2
we arrive at the identity

) sin p 1 ( ) sin m 1 (
) sin k 1 (
2 2 2 2
3 2 2

+
+
sin m 1
1
) p m ( m
) p m 2 pk m k 2 m ( ) m k (
p m
k
2 2 2 2 4
2 2 2 2 4 2 2 2
4
6
sin p 1
1
) p m ( p
) p k (
sin m 1
1
m ) p m ( m
) m k (
2 2 2
3 2
2 2 2 2 2
3 2 2

,
_

(J.4)
Appendix K. Elliptic Integrals and their Series Expansions.
The Elliptic Integral of the First Kind, denoted by ) m , a ( F , is defined by

a
0
2 2
sin m 1
d
) m , a ( F
and has the power series expansion

0 n
n 2
n 2
) a ( S m
) 2 / 1 ( I ! n
)
2
1
n ( I
) m , a ( F
where
59 Page

1
1
]
1

,
_

,
_

1 p
n 2
n 2
p
pa 2 sin
p n
n 2
a
n
n 2
2
1
) a ( S
The complete Elliptic Integral of the First Kind, denoted by ) m ,
2
a
( F ) m ( K is
defined by

2 / a
0
2 2
sin m 1
d
) m ,
2
a
( F ) m ( K
and has the series expansion

1
1
1
]
1

,
_

+ +

,
_

,
_

+
n 2
2
2
4
2
2
2
m
! n 2
! ! ) 1 n 2 (
m
4 2
3 1
m
2
1
1
2
a
) m ( K
The Elliptic Integral of the Third Kind, denoted by ) m ; n , a ( H is defined by

a
0
2 2 2
sin m 1 ) sin n 1 (
d
) m ; n , a ( H
For this integral we have the expansions
(i) For 1 | n | < ,

0 p
n 2
) 2 / 1 (
p
p
) a ( S ) x ( B n ) k ; n , a ( H
where

,
_

p
0 m
m ) 2 / 1 (
p
x
m
2 / 1
) x ( B ,
n
k
x
2

2 m
m
) ! m ( 4
! ) m 2 ( ) 1 (
m
2 / 1

,
_

(ii) For 1 n >

60 Page
+

'

) k ; m , a ( H
k m
m ) m 1 (
m n
1
) k ; n , a ( H
2

'

1
1
]
1

+
+ +

2
1
2
k n 1
k 2 sin
2
n m
tan ) k , a ( F
nm
k
m n
1
where

n 1
k n
m
2
+
+
( 1 m > )
Appendix L. The Weirstrass and Jacobi Elliptic Functions.
The relation between Weirstrass and Jacobi elliptic functions is given by

) m ; x ( sn
e e
e ) x (
2
3 1
3

+
where

3 1
e e
Using the above relation, the reader can easily establish the equivalence between
the solutions expressed in terms of Weirstrass elliptic functions and those
expressed in terms of Jacobi elliptic functions.
We also have the following expansions:
+ + + +
! 5
x
) m m 14 1 (
! 3
x
) m 1 ( x ) m ; x ( sn
5
4 2
3
2
and

+
2 k
2 k 2
k
2
x c
x
1
) x (
where

20
g
c
2
2
,
28
g
c
3
3
and

2 k
2 m
m k m k
c c
) 1 k 2 )( 3 k (
3
c ( 4 k )
61 Page
References
[1] S. Weinberg: Gravitation and Cosmology
Wiley 1972
[2] L. D. Landau and E. M. Lifshitz: The Classical Theory of Fields
Pergamon 1971
[3] C. A. Korn and T. M. Korn: Mathematical Handbook
Second Edition, McGraw-Hill 1968
[4] M. V. Berry: Principles of Cosmology and Gravitation
Institute of Physics Publishing,
Bristol and Philadelphia, 1993
[5] M. Abramowitz and I. Stegun: Handbook of Mathematical Functions
Dover 1965
[6] E. T. Whittaker and G. N. Watson: A Course of Modern Analysis
Cambridge University Press, 4
th
Edition 1927
[7] H. T. Davis: Introduction to Nonlinear Differential and Integral Equations
Dover 1962