RUSSIAN ACADEMY OF NATURAL SCIENCES

INTERNATIONAL INSTITUTE FOR THEORETICAL

AND APPLIED PHYSICS

G. I. Shipov

A THEORY
OF PHYSICAL
VACUUM
A New Paradigm

Moscow 1998
UDC 530.1

Reviewers:
Professor of Physics, State University of Moscow
R.N.Kysmin,

Professor of Physics, Physical Institute of RAN

A.A.Ruchadze.

Shipov G.I
A Theory of Physical vacuum: A New Paradigm.
This book was originally published in the Russian language by Moscow
ST-Center, Russia 1993. { 362 p. English edition { 312 p.

ISBN 5-7273-0011-8
The monograph contains the results of the programe of the geometri-
zation of the equations of physics advanced by Einstein at the beginning
of the 20th century and developed by the author who gives the idea of the
universal principle of relativity and the theory of physical vacuum and
analyses theoretical and experimental eects of the theory.
The second part of the monograph develops the mathematical system
of the physical vacuum theory. It contains the main characteristics of the
absolute parallelism geometry in vector and spinor basis.
The book is intended for specialists in theoretical physics, teachers,
post-graduates, students and everyone interested in new physical theo-
ries.
1604030000 02 Without announcement
5 T O (03) 93

ISBN 5-7273-0011-8 c G.I.Shipov, 1998

c Intrnational Institute
for Theoretical and Applied
physics RANS, 1998
c Translated into
English by A.P.Repiev, 1998
 3

Preface
This book gives a concise presentation of ideas and methods used by the
author to develop the Cliord-Einstein program of the geometrization of the
equations of physics, and also to solve various fundamental problems of modern
theoretical physics proceeding from the concept of the universal principle of
relativity and the theory of physical vacuum. In his studies the author made
an attempt to combine phenomena of seemingly dierent nature and to sketch
a coherent picture of modern physics.
The author is most grateful to V. Yu. Tatur and all those who, directly or
indirectly, made the publication of this book possible. Special thanks are due
to my friends and colleagues E. A. Gubarev, A. N. Sidorov, and I. A. Volodin.
Many ideas expounded in this book were presented in my rst monograph
published in 1979 with a support of M. A. Adamenko and I. S. Lakoba at
Moscow University Press.
I remember with gratitude my productive talks with V. Skalsky, an Associate
Professor at Slovak Polytechnic, who made some valuable points about various
vacuum states of matter.
Useful comments of A. E. Akimov have been encouraging in many respects
for my studies of torsion elds and interactions.
The attention and support of all these persons contributed enormously to
the publication of this book.
Last but not least, the author must record his deep obligation to Elena
Turantaeva who was good enough to edit the book and read the proofs.
1993
Gennady Shipov

Preface to English edition

The translation of the book into English is shortened as far as it doesn't
contain the 5th chapter of the Russian edition. This chapter is dedicated to
phenomena of seemingly dierent nature and its absence doesn't in uence the
main scientic results.

1998
Gennady Shipov
4

Conventions 1

Three-dimensional tensor indices are denoted by the Greek letters ; ; ; : : :

and take the values 1, 2, 3.
Three-dimensional vectors (e.g., linear and angular velocity) are denoted as:
~v and ~! or v and !:
Four-dimensional tensor indices are denoted by Latin letters i; j; k : : :; they
assume the values 0, 1, 2, 3. Letters from the rst part of the alphabet
(a; b; :: :; h) are used as tetrad indices, e.g., ei a ; (a = 0; 1; 2; 3):
Spinor indices in the spinor
-basis are denoted by Roman capitals A; B : : :; C_ : : : D_
and take the values 0,1 or 0_ ; 1_ . Spinor indices in the -basis are labeled by Greek
_ Æ_ : : ::
letters ; ; : : : ; ;
Symmetrization and antisymmetrization of pairs of indices:

S(ij ) = 1 (Sij + Sji); S[ij ] = (Sij Sji ):

1
2 2
Exclusion of an index from symmetrization or antisymmetrization:

S(ijj jk) = 1 (Sijk + Skji ); S[ijj jk] = 1 (Sijk Skji ):

2 2
Passing over to local (tetrad) indices: S abc = ea i S ijk ej b ek c.
External product: ea ^ ec = aa ec ec ea .

The Levi-Chivita pseudotensor: "ijkm ; dual tensor: S ij = 12 "ijkm S km .
The matrix representation of
(a) tensor quantities :
S a bk or, discarding the matrix indices a and b, S abk ! Sk ;
(b) spin-tensor quantities: S ABC_ Dk
_ ! Sk .
A matrix product: [Tm; Tk ] = Tm Tk Tk Tm .
Hermitian conjugate matrices: S +B Dkn _ .

Derivatives

Partial derivatives with respect to the translational coordinates xi are labeled

by a comma in front of an index, i.e., f;k = @f=@xk = @k f ; a covariant derivative
with respect to the Christoel symbols i jk is denoted by rk or rk ui = @k ui +
i uj .
jk
A local covariant derivative: raub = @aub + bca uc .

A covariant derivative rk with respect to the connection
ijk = eia eaj;k of

the A4 geometry: rk ui = @k ui +
ijk uj .
1 The following is a list of only some important notations. All the conventions are explained
in the text.
 5

An external derivative: d.
A spinor derivative: @AB_ .

Translational metric and tetrads

Translational coordinates: x0 ; x1 ; x2 ; x3 .
The metric signature: (+ ):
A translational linear element:
ds2 = abeai ebj dxi dxj , ab =  ab = diag(1 1 1 1)
The structural equations of the group of translations of the A4 geometry:
r[arb] xi =
ab: : c rc xi .
1-form of the tetrad: ea = eai dxi .

Rotational metric and torsion

Rotational coordinates: '1 ; '2 ; '3 ; 1 ; 2 ; 3 :

Rotational metric: d 2 = db adab = T abiT baj dxi dxj ,
dab = dab :
The torsion of A4 geometry:
jk : :i = ei ea 1 i a a
a [k;j ] = 2 e a (e k;j e j;k ):
The contorsion tensor of A4 geometry (the rotational Ricci coeÆcients):
: :s + g
: :s ) = ei r ea :
Tjki =
jk: :i + g im(gjs
mk ks mj a k j

1-form of contorsion: T ab = T abk dxk = T abc ec ; T(ab) = 0:

The structural equations of a rotational group (the matrix indices are dis-
carded): r[k rm] ei = 12 Rkm ei , where Rkm = 2r[mTk] + [Tm; Tk ]:

Connection and curvature of A4 geometry

Connection:
ijk = i jk + T ijk = ei a eaj;k ;

i[jk]
= T[ijk] =
jk
: :i ;
i = i + g im(g
:: s + g
: :s ):
(jk) jk js mk ks mj
Curvature:
i
Sjkm = 2
ij [m;k] + 2
is[k
sjj jm] =
= Rijkm + 2r[k Tjijjm] + 2Tci[k Tjcj jm] = 0

where Rijkm = 2 ij [m;k] + 2 is[k sjj jm] | the Riemann tensor.

1-form of connection:
ab =
abk dxk =
abc ec :
The Cartan structural equations:
(a) rst structural equations: dea ec ^ T ac = 0;
(b) second structural equations: Rab + dT ab + T cb ^ T ac = 0:

Spinor
-basis
Newman-Penrose symbols: iAB_ :
6

Translational metric: gij = "AC "B_ D_ iAB_ jC D_ , where "AB is a fundamental
spinor  
AB C_D 0 1
" = "AB = " = "C_ D_ =
_
1 0 :
The rotational Ricci coeÆcients:

TABC
_ Dk
_ = Ci D_ rk ABi
_ :

The rotational Ricci coeÆcients in terms of Carmeli matrices: TAB_ with

matrix elements (TAB_ )C D :
The Riemann curvature in terms of Carmeli matrices: RABC_ D_:
The equations of physical vacuum written in Carmeli matrices:

@C D_ Ai B_ @AB_ Ci D_ = (TC D_ )AP Pi B_ + Ai B_ (TDC R_

_ ) B
+
_

(TAB_ )C P Pi D_ Ci R_ (TBA R_ ;

_ ) D
+
_ (As)
RF ED
_ B_ = @DB
_ TF E
_ @F E_ TDB_ (TDB_ )F S TS B_ (TED F_
_ ) B_ TF F_ +
+

+(TF E_ )D S TS B_ + (TEF F_
_ ) B_ TDF_ + [TF E
+
_ ; TDB
_ ]: (B s+ )
+ Hermitially conjugate equations.
 Part II
GEOMETRY
OF ABSOLUTE
PARALLELISM
 9

Introduction
Geometry with absolute parallelism was rst considered in 1923-24 in the
works of Weitzenbock [1, 2] and Vitali [3, 4]. Weitzenbock suggested that there
exist in the n-dimensional manifold M with coordinates x1 ; : : : ; xn of Rieman-
nian spaces with a zero Riemann-Christoel tensor
S ijkm = 2
ij [m;k] + 2
is[k
sjj jm] = 0: (4.1)
Relationship (4.1) was regarded as the condition of parallel displacement of
an arbitrary vector in a given space in the absolute (independent of path) sense.
In 1924 Vitali introduced the concepts of the connection of absolute parallelism
[3]

kij = eka eai;j; (4.2)
;j = @
@xj ; i; j; k : : : = 0; 1; 2; 3;
a; b;c : : : = 0; 1; 2; 3;
where eka and eai are basic vectors dened at each point of space and translatable
in the absolute sense to any point of the space in any direction. Weitzenbock
[5] showed that the connection (4.2) can be represented as the sum

ijk = ijk + Tjk
i ; (4.3)
where
i 1
= g im(gjm;k + gkm;j
jk gjk;m); (4.4)
2
are the Christoel symbols and
Tjki =
::i im s ::s
jk + g (gjs
mk + gks
mj ) (4.5)
are the Ricci rotation coeÆcients [6] for the basis eai .
The tensor
::i
jk , dened as
1 i a

::i i a a
jk = e a e[k;j ] = 2 e a (e k;j e j;k ); (4.6)

came to be known as the anholonomity object [7], therefore the emergence of

the geometry of absolute parallelism continued the development of anholonomic
dierential geometry [8].
Cartan and Schouten [9, 10], proceeding from the group properties of the
space of constant curvature, introduced the connection (4.3), in which the com-
ponents of the Ricci rotation coeÆcients (4.5) are constants.
Cartan and Schouten reasoned as follows. Suppose that in a n - dimensional
dierentiable manifold M with the coordinates x1 ; : : :; xn we have n contrava-
riant vector elds
aj = aj (xk ); (4.7)
where
a; b; c : :: = 1 : : : n
10 PART 2.

are vector indices, and

i; j; k : : : = 1 : : :n
are coordinate indices.
Suppose that
det(aj ) 6= 0
and that the functions aj satisfy the equations
aj  kb;j ::f  k ;
bi  ka;j = Cab f
::f have the following properties:
where the constants Cab
::f = C ::f ;
Cab (4.8)
ba
::a C ::f + C ::a C ::f + C ::a C ::f = 0:
Cfb (4.9)
cd fc db fd bc
We can then say that we have an n-parametric simple transitive group (group
::f are structural constants of the
Tn ) operating in the manifold such that Cab
group that obey the Jacobi identity (4.9). The vector eld bj is said to be
innitesimal generators of the group.
Let now the basis ekb , dened at each point of the manifold M , meet the
condition
det(eja ) 6= 0:
If we suppose that
eja (xk0 ) = aj (xk0 );
where xk0 are the coordinates of some arbitrary point P , then we have for the
function eja (xk0 ) the equations
ejaekb;j ejb eka;j = Cab
::f ek :
f (4.10)
It follows from the normalization condition for the basis
eai eja = Æij ; eaiei b = Æba; (4.11)
and from (4.10), that
::i = 2ei ea
Cjk i ::a b c
a [k;j ] = e aC bc ej ek : (4.12)
Comparing (4.8) and (4.6), we see that
1 ::i

::i
jk = 2 Cjk ;

i.e., the components of the anholonomity object of a homogeneous space of

absolute parallelism are constant.
It is easily seen that the connection (4.2) possesses a torsion. In our specic
case
1 ::i

k[ij ] =
::k k
ij = T[ij ] = 2 Cjk :
INTRODUCTION 11

It was exactly in this manner that Cartan and Schouten introduced connection
with torsion [9, 10]. Therefore, the development of the geometry of absolute
parallelism brought about the emergence of the Riemann-Cartan geometry with
the connection
1

ijk = ijk + (Cijk Cjki Ckij ); (4.13)
2
where Sijk = 12 Cijk is the torsion of space.
Further development of the geometry of absolute parallelismin the n-dimensional
dierentiable manifold M with coordinates x1 ; : : : ; xn (geometries An ) is de-
scribed in the works of Bortolotti [11{14], Griss [15], Schouten [16, 17], Eisen-
hart [18] and other authors [19-25]. Specically Bortolotti [12] was the rst to
point out that the Cartan-Schouten connection and the Weinzbock-Vitali (4.2)
connection is one and the same thing. Besides, Bortolotti showed that the tensor
(4.1) can be represented as the sum
S ijkm = Ri jkm + 2r[k Tjij jm] + 2Tci[k Tjcjjm] = 0; (4.14)
where
Ri jkm = 2 i i s
j [m;k] + 2 s[k jj jm] (4.15)
is the Riemann tensor, and the i are given
quantities Tjk by (4.5).
In 1937 Thomas [20, 21] approached absolute parallelism as parallel displa-
cement of vectors "in toto," since the connection of space An (just as that of
a at space En) is integratble. Therefore, a vector specied at some point An
can be specied at any other point of space. Lastly, the works [23{25] give a
classication of spaces with absolute parallelism.
Geometry A4 has been rst used by Einstein [26] in applications to problems
of theoretical physics. The scientist made an attempt to combine the equations
of his theory with the equations of the Maxwell-Lorentz electrodynamics [27].
We note in passing that within the framework of the geometry of absolute
parallelism Einstein has written most (all in all 13) works.
By developing Einstein's program to construct a unied eld theory, this
author came to the conclusion that it is necessary to use the A4 geometry
as a geometry of space of events in universal relativity theory and the theory
of physical vacuum. Unlike Einstein and his following, the author employed
Cartan's structural equations of the geometry of absolute parallelism, which
are generalizations of Einstein's vacuum equations Rik = 0 for the case where
the energy-momentum tensor on the right-hand side of Einstein's equations is
geometric in nature.
The program of unied eld theory put forward by Einstein boils down to
solving two strategic problems of modern theoretical physics:
(a) the minimum program has it as its goal to geometrize the equations
of electromagnetic eld and to combine them with the equations of Einstein's
theory of gravitation;
(b) the maximum program is aimed at the search for completely geometrized
equations of the gravitational and electromagnetic eld (including sources), i.e.,
the geometrization of the elds that form matter.
12 PART 2.

Although much time was devoted to this search (around 30 years), Einstein
failed to solve the problem in a form acceptable to science. Together with many
outstanding scientists of the time he wrote a wealth of works relying on various
geometries. But all of them failed to meet the above requirements (a) and
(b). Also, it was unclear how to geometrize spin elds (e.g., Dirac's eld) that
are sources of electromagnetic elds. Wheeler added to the program of unied
eld theory a further point that required a spinor treatment of the equations of
the unied eld. The latter condition can be met in the case where the main
geometric quantities of the theory are spinors rather than tensors. A spinor
treatment of classical geometries was given in the works by Penrose [38, 40, 54],
which was of much help to me in my constructing a theory of physical vacuum,
a present-day outgrowth of Einstein's program of unied eld theory.
Chapter 5

Geometry of absolute
parallelism in vector basis
5.1 Object of anholonomicity. Connection of
absolute parallelism
Consider a four-dimensional dierentiable manifold with coordinates xi (i =
0; 1; 2; 3) such that at each point of the manifold we have a vector eai (i =
0; 1; 2; 3) and a covector ejb (b = 0; 1; 2; 3) with the normalization conditions
eaieja = Æij ; eai ei b = Æba : (5.1)
For arbitrary coordinate transformations
0
0 @xi
dxi = k dxk (5.2)
@x
in coordinate index i the tetrad eai transforms as a vector
i
eai0 = @xi0 eai: (5.3)
@x
In the process, in the tetrad index a relative to the transformations (5.2) it
behaves as a scalar.
Tetrad eai denes the metric tensor of a space of absolute parallelism
gik = abeai ebk ; ab =  ab = diag(1 1 1 1) (5.4)
and the Riemannian metric
ds2 = gik dxi dxk : (5.5)
Using the tensor (5.4) and the normal rule [29], we can construct the Christof-
fel symbols
i = 1 g im(g
jk jm;k + gkm;j gjk;m): (5.6)
2
13
14 CHAPTER 5. GEOMETRY OF ABSOLUTE. . .

that transform following a nontensor law of transformation [29]

k00 0 @ 2xk @xk0 @xi @xj @xk0 k

=
@xi0 @xj 0 @xk + @xi0 @xj 0 @xk
ji ji (5.7)

with respect to the coordinate transformations (5.2). In the relationship (5.6)

and farther on we will denote the partial derivative with respect to the coordi-
nates xi as
;k = @ : (5.8)
@xk
Dierentiating the arbitrary vector eai gives
@xj
eai;j0 = @xj 0 eai;j : (5.9)

Applying the dierentiation operation (5.9) to the relationship (5.3) gives

@xi @xj @ 2xi
eai0 ;j 0 = @xi0 @xj 0 eai;j + @xi0 @xj 0 eai : (5.10)

Alternating the indices i0 and j 0 and subtracting from (5.10) the resultant
expression, we have
@xi @xj
eai0 ;j 0 eaj 0 ;i0 = (eai;j eaj;i) @xi0 @xj 0 :
Considering (5.3), we can rewrite this relationship in the form
i j k0
eaj 0;i0 ) = eka (eai;j eaj;i) @xi0 @xj 0 @x k :
0
eka(eai0 ;j 0
@x @x @x
By denition, the dierential
dsa = eai dxi (5.11)
is said to be complete, if the following relationship holds:
eai;j eaj;i = 0: (5.12)
Otherwise, for eai;j eaj;i 6= 0, the dierential (5.11) is not integrable (equal-
ity (5.12) is the condition of integration for the relationship (5.11)).
We will introduce the following geometric object [30]
1 i a

::i i a a
jk = e a e[k;j ] = 2 e a (e k;j e j;k ) (5.13)

with a tensor law of transformation relative to the coordinate transformations

(5.2)
0
0 @xj @xk @xi

::i
j 0 k0 =
::i
jk @xj 0 : (5.14)
@xk0 @xi
5.2. COVARIANT DIFFERENTIATION. . . 15

Clearly, if the condition (5.12) is met, this object vanishes. In that case,
tetrad eai is holonomic and the metric (5.5) characterizes holonomic dierential
geometry. If the object (5.13) is nonzero, we deal with anholonomic dierential
geometry, and the object (5.13) itself is called an object of anholonomicity.
We will rewrite the relationship (5.10) in the following manner:
@ 2 xi ea + @xi @xj ea =
eai0 ;j 0 =
@xi0 @xj 0 i @xi0 @xj 0 i;j

@ 2 xk + @xi @xj
k ea ;
= (5.15)
@xi0 @xj 0 @xi0 @xj 0 ij k
where we have introduced the notation

kij = ekaeai;j (5.16)
and used the orthogonality condition (5.1).
It is seen from the relationships (5.15 ) that the object
kij gets transformed
relative to the transformations (5.2) as the connection
0 0
0 @ 2 xk @xk + @xi @xj @xk
k :

ki0 j 0 = (5.17)
@xi0 @xj 0 @xk @xi0 @xj 0 @xk ij
The connection of a space given by (5.16) is called the connection of absolute
parallelism [31].
Interchanging in (5.17) the indices i and j gives
0 @ 2 k
x @x k0 @x i @xj @xk0

kj 0i0 = j 0 i0 k + j 0 i0 k
kji: (5.18)
@x @x @x @x @x @x
Subtracting (5.18) from (5.17) gives
0 @xi @xj @xk0 k

k[i0j 0 ] = j 0 i0 k
[ij ]: (5.19)
@x @x @x
It follows from the relationships (5.16) and (5.13) that the connection of
absolute parallelism features the torsion

k[ij] =
::k
ij ; (5.20)
dened by the object of anholonomity.

5.2 Covariant dierentiation in A4 geometry. Ricci

rotation coeÆcients
The denition of the covariant derivative with respect to the connection of the
geometry of absolute parallelism (A4 geometry)
ijk from a tensor of arbitrary
i:::p has the form
valence Um:::n
 i:::p = U i:::p +
i U j:::p + : : : +
p U i:::j
rk Um:::n m:::n;k jk m:::n jk m:::n (5.21)

jmk Uj:::n
i:::p : : :
j U i:::p :
nk m:::j
16 CHAPTER 5. GEOMETRY OF ABSOLUTE. . .

This denition enables some quite useful relationships in A4 geometry to be

proved.
Proposition 5.1. Parallel displacement of the tetrad eai relative to the
connection
ijk equals zero identically.
Proof. From the denition (5.21) we have the following equalities:

rk ei a = aia;k +
ijk eja; (5.22)

rk eaj = eaj;k
ijk eai : (5.23)
Since the connection
ijk is dened as

ijk = ei a eaj;k ; (5.24)
we have
ei aeaj;k
ijk = 0:
Multiplying this equality by eai and taking into consideration the orthogo-
nality conditions (5:1), we get

rk eaj = eaj;k
ijk eai = 0: (5.25)
To prove that the relationship (5.22) is zero, we will take a derivative of the
convolution eaj ei a = Æji
(Æji );k = (eaj eia);k = ei a eaj;k + eaj eia;k = 0:
Hence, by (5.24), we have

ijk = eaj ei a;k (5.26)
or
eaj eia;k +
ijk = 0:
Multiplying this relationship by eja and using the conditions eaj ei a = Æji , we
have 
rk ei a = eia;k +
ijk eja = 0: (5.27)

Proposition 5.2. Connection
ijk can be represented as the sum

ijk = ijk + Tjk
i ; (5.28)
where ijk are the Christoel symbols given by the relationship (5.6), and
Tjki =
::i im ::s ::s
jk + g (gjs
mk + gks
mj ) (5.29)
are the Ricci rotation coeÆcients [30].
Proof. Let us represent the connection (5.28) as the sum of parts symmetrical
and skew-symmetrical in indices j; k

ijk =
i(jk) +
i[jk]; (5.30)
5.2. COVARIANT DIFFERENTIATION. . . 17

where
1 1

i(jk) = (
ijk +
ikj );
i[jk] = (
ijk
ijk ):
2 2
We now add to and subtract from the right-hand side of (5.30) the same ex-
pression

ijk =
i(jk) +
i[jk] + g im(gjs
s[km] + gks
s[jm])
(5.31)
g im(gjs
s[km] + gks
s[jm]):
We then group the terms on the right-hand side of (5.31) as follows:

ijk =
i(jk) g im(gjs
s[km] + gks
s[jm]) +
+
i[jk] + g im(gjs
s[km] + gks
s[jm]): (5.32)
Since

i[jk] =
::i
jk ;
it follows from (5.32) and (5.29) that

ijk =
i(jk) g im(gjs
s[km] + gks
s[jm]) + Tjki : (5.33)
We now show that
i =
i(jk) g im(gjs
s[km] + gks
s[jm]): (5.34)
jk
Actually, we have the relationships
1

i(jk) = ei a ea(j;k) = ei a(eaj;k + eak;j );
2
1

i[jk] = ei a ea[j;k] = ei a (eaj;k eak;j );
2
gjs = ab eaj ebs; (5.35)
therefore (5.34) become
i
= eia ea(j;k) + g im(ab eaj eb[m;k] + ab eak eb[m;j ]) =
jk
1
=  cd ab eic em b a b c
d (em e j;k + em e k;j ) +
2
1

+ g im ab (eaj ebm;k eaj ebk;m ) + ab(eak ebm;j eak ebj;m) :
2
Regrouping the terms here gives
i 1

jk = g im (abeaj ebm );k + (ab eak ebm );j (ab eaj ebk );m :
2
Hence, by (5.35), we obtain
i 1
jk = g im(gjm;k + gkm;j gjk;m); (5.36)
2
18 CHAPTER 5. GEOMETRY OF ABSOLUTE. . .

or
1 im
g (gjm;k + gkm;j gjk;m ) =
2
=
i(jk) g im (gjs
s[km] + gks
s[jm]) = ijk : (5.37)
Substituting (5.37) into (5.33), we get the relationship (5.28).
Proposition 5.3. The Ricci rotation coeÆcients Tjk i can be represented
in the form
Tjki = ei ark eaj ; (5.38)
Tjki = eaj rk ei a ; (5.39)
where rk stands for a covariant derivative with respect to the Christoel ijk
symbols.
Proof. We will represent in the relationships (5.25) and (5.27) the connection

ijk as the sum (5.28)

rk eaj = eaj;k i ea
jk i Tjki eai = 0; (5.40)

rk ei a = eia;k + i ej + T i ej
jk a jk a =0 (5.41)
Since, by denition [29], we can write
rk eaj = eaj;k i ea ;
jk i

rk ei a = eia;k + i ej ;
jk a
then (5.40) and (5.41) can be written as
rk eaj Tjki eai = 0; (5.42)

rk ei a + Tjki eja = 0: (5.43)

Multiplying (5.42) by ei a and (5.43) by eaj , respectively, we will obtain
(using the orthogonality conditions (5.1)), by (5.42), (5.43), the relationships
(5.38) and (5.39).

We will now calculate the covariant derivative rk with respect to the metric
tensor g jm, knowing that g jm =  ab eja emb
  
rk g jm =rk  ab eja emb =rk eja ema =
 
= ema rk eja + eja rk ema :
From the relationships (5.25) and (5.27), we have

r g jm = 0: (5.44)
5.3. CURVATURE TENSOR. . . 19

On the other hand, applying the formula (5.21) to the relationship (5.44),
we nd that 
rk g jm = g;kjm +
jpk g pm +
mpk g jp = 0: (5.45)
Substituting the connection
ijk as the sum (5.28), we will write the rela-
tionship (5.45) in the form

rk g jm = rk g jm + Tpkj g pm + Tpkm g jp = 0: (5.46)
From the equality
rk g jm = g;kjm + j g pm + m g jp
pk pk = 0; (5.47)
we have, by (5.46),
Tpkj g pm + Tpkmg jp = Tkjm + Tkmj = 0:
This equality establishes the following symmetry properties for the Ricci rota-
tion coeÆcients:
Tjmk = Tmjk : (5.48)
Therefore, in the A4 geometry the Ricci rotation coeÆcients have 24 indepen-
dent components.

5.3 Curvature tensor of A4 space

The curvature tensor of the space of absolute parallelism S ijkm is dened in
terms of the connection
ijk following a conventional rule [18]
S ijkm = 2
ij [m;k] + 2
is[k
sjj jm] = 0; (5.49)
where the parentheses [ ] signify alternation in appropriate indices, whereas the
index within the vertical lines j j is not subject to alternation.
Proposition 5.4. The Riemann-Christoel tensor of a space with the
connection (5.26) equals zero identically.
Proof. From the relationship (5.26) we have
eaj;k =
ijk eai : (5.50)
Dierentiating the relationship (5.50) with respect to m gives
eaj;k;m = (
ijk eai );m =
ijk;m eai + eai;m
ijk =
= (
ijk;m + ei a eas;m
sjk )eai = (
ijk;m +
ism
sjk )eai:
Alternating this relationship in indices k and m we get
2eaj;[k;m] = 2(
ij[m;k] + 2
is[k
sjj jm]) = S ijkm eai : (5.51)
20 CHAPTER 5. GEOMETRY OF ABSOLUTE. . .

Since the operation of dierentiating with respect to indices k and m is

symmetrical, we have
eaj;[k;m] = 0;
From this equality, considering that eai in (5.51) is arbitrary, we will get
S ijkm = 0: (5.52)

Proposition 5.5. Tensor S ijkm can be represented as the sum

S ijkm = Ri jkm + 2r[k Tjijjm] + 2Tci[k Tjcj jm] = 0; (5.53)
where
Rijkm = 2 i i s
j [m;k] + 2 s[k jj jm] (5.54)
is the tensor of the Riemannian space A4 .
Proof. Substituting the sum
ijk = ijk + Tjk
i into (5.49) gives

S ijkm = 2 i i s i i s
j [m;k] + 2 s[k jj jm] + 2Tj [m;k] + 2Ts[k Tjj jm] +
2Tsi[k sjj jm] + 2 is[k Tjsjjm] = 0: (5.55)
Using (5.54), we will write (5.55) as follows:
S ijkm = Ri jkm + 2Tji[m;k] + 2Tsi[k Tjsj jm] +
+2 sj[k Tjisjm] + 2 is[k Tjsjjm] = 0: (5.56)
If now we add to the right-hand side of this relationship the expression
2 s[km] Tsji = 0;
and take into consideration that [29]
i:::p = U i:::p +
rk Um:::n i U j:::p + : : : + p U i:::j
m:::n;k jk m:::n jk m:::n
j U i:::p : : : j U i:::p ; (5.57)
mk j:::n nk m:::j
we will obtain from (5.56) the equality (5.53).
Let us now rewrite the relationship (5.53) as
Ri jkm = 2Tji[m;k] 2Tsi[k Tjsj jm]: (5.58)
Substituting here (5.38) and (5.39)
Tjki = ei ark eaj ; Tjki = eaj rk eia ;
we obtain
2Tji[m;k] = 2ei ar[k rm] eaj 2r[k eijajrm] eaj ;
2Tsi[k Tjsj jm] = 2easr[k eija esajrm] eaj = 2r[k eijaj rm]eaj :
5.3. CURVATURE TENSOR. . . 21

Therefore, it follows from the relationships (5.58) that

Ri jkm = 2ei a r[krm] eaj = 2eia r[mrk] eaj : (5.59)

Proposition 5.6. The torsion eld

::i
jk of the A4 space satises the
equations

r k
::i
[jm+ 2
::s
]
::i
[kj
m]s = 0: (5.60)
Proof. Alternating the expression (5.49) in indices j; k; m and using the rela-
tionship
i[jk] =
::i
jk , we get
S[ijkm] = 2
::i i ::s
[jm;k] + 2
s[k
jm] = 0: (5.61)
If then we add and subtract here the quantity
2
s[kj
::i s ::i
jsjm] + 2
[km
j ]s;
we will have
2
::i i ::s
[jm;k] + 2
s[k
jm] 2
s[kj
::i s ::i
jsjm] 2
[km
j ]s +
2
s[kj
::i s ::i
jsjm] + 2
[km
j ]s = 0:
Using the formula (5.21), we can rewrite this relationship as follows:

2 r[k
::i ::s ::i ::s ::i
jm] 2
[kj
jsjm] 2
[km
j ]s =
 (5.62)
= 2 r[k
::i ::s ::i
jm] + 4
[kj
m]s = 0;
whence we have (5.60).
Proposition 5.7. The Riemann tensor Ri jkm of the A4 space satises
the equality
Ri[jkm] = 0: (5.63)
Proof. Alternating the relationship (5.54) in indices j; k; m and using the equal-
ity
T[ijk] =
::i
jk ;
we have
Ri[jkm] = 2r[k
::i i ::s
jm] + 2Ts[k
jm]:
If in the right-hand side of the equality we add and subtract the quantity
2T[skj
::i s ::i
jsjm] + 2T[km
j ]s;
we obtain
Ri[jkm] = 2r[k
::i i ::s s ::i s ::i
jm] + 2Ts[k
jm] 2T[kj
jsjm] 2T[km
j ]s +

+2T[skj
::i s ::i ::i ::s ::i
jsjm] + 2T[km
j ]s = 2 r[k
jm] 2
[kj
jsjm]

2
::s ::i ::i ::s ::i
[km
j ]s = 2 r[k
jm] + 4
[kj
m]s = 0;

which proves the validity of the relationship (5.63).

22 CHAPTER 5. GEOMETRY OF ABSOLUTE. . .

5.4 Formalism of external forms and the

matrix treatment of Cartan's structural
equations of the absolute parallelism
geometry
Consider the dierentials
dxi = eaei a ; (5.64)
dei b =
ab eia ; (5.65)
where
ea = eai dxi ; (5.66)
a a i a

b = e i de b =
bk dx k (5.67)
are dierential 1-forms of tetrad eai and connection of absolute parallelism
abk .
Dierentiating the relationships (5.64), (5.65) externally [31], we have, respec-
tively,
d(dxi ) = (dea ec ^
ac )ei a = S a ei a ; (5.68)
d(dei a) = (d
ba
ca ^
bc )ei b = S baei b : (5.69)
Here S a denotes the 2-form of Cartanian torsion [31], and S ba { the 2-form
of the curvature tensor. The sign ^ signies external product, e.g,
ea ^ eb = ea eb eb ea : (5.70)
By denition, a space has a geometry of absolute parallelism, if the 2-form
of Cartanian torsion S a and the 2-form of the Riemann-Christoel curvature
S ba of this space vanishe
S a = 0; (5.71)
S ba = 0: (5.72)
At the same time, these equalities are the integration conditions for the
dierentials (5.64) and (5.65).
Equations
dea ec ^
ac = S a ; (5.73)
d
ba
ca ^
bc = S ba ; (5.74)
which follow from (5.68) and (5.69), are Cartan's structural equations for an
appropriate geometry. For the geometry of absolute parallelism hold the condi-
tions (5.71) and (5.72), therefore Cartan's structural equations for A4 geometry
have the form
dea ec ^
ac = 0; (5.75)
d
ba
ca ^
bc = 0: (5.76)
5.4. FORMALISM OF EXTERNAL FORMS. . . 23

Considering (5.28), we will represent 1-form
ab as the sum

ab = ab + T ab : (5.77)
Substituting this relationship into (5.75) and noting that
ec ^
ac = ec ^ T ac ;
we get the rst of Cartan's structural equations for A4 space.
dea ec ^ T ac = 0: (A)
Substituting (5.77) into (5.76) gives the second of Cartan's equations for A4
space.
Rab + dT ab T cb ^ T ac = 0; (B )
where Rab is the 2-form of the Riemann tensor
Rab = d b^
a c a: (5.78)
b c

By denition [31], we always have the relationships

dd(dxi ) = 0; (5.79)
dd(dei a ) = 0: (5.80)
In the geometry of absolute parallelism these equalities become
d(dea ec ^ T ac ) = Racfdec ^ ef ^ ed = 0; (5.81)

d(Rab + dT ab T cb ^ T ac ) = dRab + Rfb ^ T af T fb ^ Raf = 0: (5.82)

Here
Racfd = 2T ac[d;f ] 2T ab[f T bjcjd]:
Equalities (5.81) and (5.82) represent the rst and second of Bianchi's iden-
tities, respectively, for A4 space. Dropping the indices, we can write Cartan's
structural equations and Bianchi's identities for the A4 geometry as

de e ^ T = 0; (A)
R + dT T ^ T = 0; (B )
R ^ e ^ e ^ e = 0; (C )
dR + R ^ T T ^ R = 0: (D )

Proposition 5.8. The matrix treatment of the rst of Cartan's struc-

tural equations (A) of the A4 geometry has the form
r k eam
[ ] eb[k T ajbjm] = 0: (5.83)
24 CHAPTER 5. GEOMETRY OF ABSOLUTE. . .

Proof. Let us write equations (A) as

dea ec ^ T ac = 0: (5.84)
Further, by (5.66), we have

dea = d(eam dxm ) = rk eam dxk ^ dxm = 1 (rk eam rmeak )dxk ^ dxm
2
and, also,

eb ^ T ab = ebk T abm dxk ^ dxm = 1 (ebk T abm ebm T abk )dxk ^ dxm :
2
Substituting these relationships into equations (5.84) we will derive the mat-
rix equations in the form
r k eam
[ ] eb[k T ajbjm] = 0; (A)
where the matrixes eam and T abm in world indices i; j; m; : : : are transformed as
vectors m
eam0 = @xm0 eam ; (5.85)
@x
@xm
T abm0 = @xm0 T abm ; (5.86)
and in the matrix indices a; b;c; : : : they are transformed as follows:
0 0
eam = aa eam ; (5.87)
0 0 0
T ab0 k = aa T abk bb0 + aa ab0 ;k : (5.88)
In relationships (5.87) and (5.88) the matrices @x =@xm form a translationm0
group T4 that is dened i
a 0 on a manifold of world coordinates x . On the other
hand, the matrices  a form a group of four-dimensional rotations O (3:1)
0
aa 2 O (3:1);
dened on the manifold of "angular coordinates" ea i . Actually, the tetrad ea i
is a mathematical image of an arbitrarily accelerated four-dimensional reference
frame. Such a frame has ten degrees of freedom: four translational ones con-
nected with the motion of its origin, and six angular ones describing variations
of its orientation. The six independent components of the tetrad ea i represent
six direction cosines of six independent angles dening the orientation of the
tetrad in space.
Proposition 5.9. The matrix rendering of the second of Cartan's struc-
turing equations (B ) of the A4 geometry has the form
Rabkm + 2r[k Tjabjm] + 2T ac[k T cjbjm] = 0: (5.89)
5.4. FORMALISM OF EXTERNAL FORMS. . . 25

Proof. We will expand the 2-form Ra d as

1 1
Rab = 2 Rabcd ec ^ ed = 2 Rabkm dxk ^ dxm : (5.90)
Further, we have
dT ab = d(T abm dxm ) = rk T abm dxk ^ dxm =
1
(r T a rm T abk )dxk ^ dxm ; (5.91)
2 k bm
and also
T ac ^ T cb = T ack T cbm dxk ^ dxm =
1
= (T ack T cbm T cbm T ack )dxk ^ dxm : (5.92)
2
Let us substitute the relationships (5.92){(5.94) into
Rab + dT ab T cb ^ T ac = 0:
Simple transformations yield
1 a
(R + rk T abm rm T abk + T ack T cbm T cbm T ack )dxk ^ dxm = 0:
2 bkm
Since here the factor dxk ^ dxm is arbitrary, we have
Rabkm + rk T abm rm T abk + T ack T cbm T cbm T ack = 0;
which is equivalent to the equations (5.89).
Proposition 5.10. The matrix form of the Bianchi identity (D ) of A4
geometry is
r[nRajbjkm] + Rcb[km T ajcjn] T cb[n Rajcjkm] = 0: (5.93)
Proof. The external dierential dRa b in the identities (D ) has the 2-form
1
dRab = 2 rn Rabkm dxn ^ dxk ^ dxm =
1
= (rnRabkm + rm Rabkn + rk Rabmn )dxn ^ dxk ^ dxm : (5.94)
6
In addition, we have
1
Rfb ^ T af = 2 Rfbkm T afndxk ^ dxm ^ dxn =
1
= (Rfbkm T afn + Rfbnk T afm + Rfbmn T afk )dxk ^ dxm ^ dxn ; (5.95)
6
1
T fb ^ Raf = T fbn Rafkm dxn ^ dxk ^ dxm =
2
1 f a
= (T bn R fkm + T bm Rafnk + T fbk Rafmn)dxn ^ dxk ^ dxm :
f (5.96)
6
26 CHAPTER 5. GEOMETRY OF ABSOLUTE. . .

Substituting relationships (5.94){(5.96) into the identity

dRab + Rfb ^ T af T fb ^ Raf = 0
and considering that dxn ^ dk ^ dxm is arbitrary, we get
rnRabkm + rm Rabkn + rk Rabmn + Rfbkm T afn + Rfbnk T afm +
+Rfbmn T afk T fbn Rafkm T fbm Rafnk T fbk Rafmn = 0;
which is equivalent to the identity (5.93).
The rst of Bianchi's identities (C ) of A4 geometry in indices of the group
O (3:1) is written as
Ra[bcd] = 0; (5.97)
or, which is the same, as
 ::f ::a
r[b
::a
cd] + 2
[bc
d]f = 0: (5.98)

5.5 A4 geometry as a group manifold.

Killing-Cartan metric
The matrix representation of Cartan's structural equations of the geometry
of absolute parallelism indicates that, in fact, this space behaves as a manifold,
on which the translations group T4 and the rotations group O (3:1) are specied.
We will consider A4 geometry as a group 10-dimensional manifold formed by
four translational coordinates xi (i = 0; 1; 2; 3) and six (by the relationship
ea i ej a = Æi j ) angular coordinates eai (a = 0; 1; 2; 3). Suppose that on this
manifold a group of four-dimensional translations T4 and a rotations group
O (3:1) are dened. We then introduce the Hayashi invariant derivative [32]
rb = ekb @k ; (5.99)
whose components are generators of the translations group T4 that is specied
on the manifold of translational coordinates xi . If then we represent as a sum
ekb = Æ kb + akb ; (5.100)
i; j; k : : : = 0; 1; 2; 3; a; b; c;: : : = 0; 1; 2; 3;
then the eld akbcan be viewed as the potential of the gauge eld of the trans-
lations group T4 [32]. In the case where akb = 0, the generators (5.99) coincide
with the generators of the translations group of the pseudo-Euclidean space E4
.
We know already that in the coordinate index k the nonholonomic tetrad
eka transforms as the vector
0
0 @xk
eka = k eka ;
@x
5.5. A4 GEOMETRY . . . 27

whence, by (5.100), we have the law of transformation for the eld aka relative
to the translationss 0
0 @xk @xk0 0
akb = n anb + n Æ nb Æ kb : (5.101)
@x @x
ei
We dene the tetrad a as
ei a = ra xi (5.102)
and write the commutational relationships for the generators (5.99) as
r arb
[ ] ab rc ;
=
::c (5.103)
where
::cab are the structural functions for the translations group of the spaceA4.
If then we apply the operator (5.103) to the manifold xi , we will arrive at the
structural equations of the group T4 of the space A4 as
r arb xi =
[ ] ab rc x

::c i (5.104)
or
r aei b
[ =
::c
]
i
ab e c : (5.105)
In this relationship the structural functions
::c ab are dened as

ab = e ir[ae b] :

::c c i (5.106)
It is seen from this equality that when the potentials of the gauge eld of
translations group akb in the relationship (5.100) vanish, so do the structural
functions (5.106). Therefore, we will refer to the eld
::c
ab as the gauge eld of
the translations group.
Considering that T c[ab] =
::c ab , we will rewrite the structural equations
(5.106) as
r[k eam] eb[k T ajbjm] = 0: (5.107)
It is easily seen that the equations (5.107) can be derived by alternating
the equations (5.42). What is more, they coincide with the structural Cartan
equations (A) of the geometry of absolute parallelism.
The structural equations of group T4 , written as (5.106), can be regarded
as a denition for the torsion of space A4 . So the torsion of space A4 coincides
with the structural function of the translations group of this space, such that
the structural functions obey the generalized Jacobi identity
 ::f ::a
r[b
::a
cd] + 2
[bc
d]f = 0; (5.108)

where rb is the covariant derivative with respect to the connection of absolute
parallelism
abc . Comparing the identity (5.108) with the Bianchi identity (5.98)
of the geometry A4 , we see that we deal with the same identity. The Jacobi
identity (5:108), which is obeyed by the structural functions of the translations
28 CHAPTER 5. GEOMETRY OF ABSOLUTE. . .

group of geometry A4 , coincides with the rst Bianchi identity of the geometry
of absolute parallelism .
The vectors
ei a = raxi ; (5.109)
that form the vector stratication [31] of the A4 geometry, point along the
tangents to each point of the manifold xi of the pseudo-Euclidean plane with
the metric tensor
ab =  ab = diag(1; 1; 1; 1): (5.110)
Therefore, the ten-dimensional manifold (four translational coordinates xi
and six "rotational" coordinates ei a) of the geometry of absolute parallelism
can be regarded as the stratication with the coordinates of the base xi and
the (anholonomic) "coordinates" of the bre ei c : If on the base xi we have the
translations group T4 , then in the bre ei c we have the rotation group O (3:1).
It follows from (5.109) that the innitesimal translations in the base xi in the
direction a are given by the vector
dsa = eaidxi : (5.111)
If from (5.111) and the covariant vector dsa = ei a dxi we form the invariant
convolution ds2 , we will obtain the Riemannian metric of A4 space
ds2 = gik dxi dxk (5.112)
with the metric tensor
gik = abeai ebk :
Therefore, the Riemannian metric (5.112) can be viewed as the metric de-
ned on the translations group T4 .
Since in the bre we have the "angular coordinates" ei a that form a manifold
in which group O (3:1) is dened, then it would be natural to dene the structural
equations for this group, as well as the metric specied on the group O (3:1).
Let us rewrite the relationships (5.38) and (5.39) in matrix form
T abk = eai T ijk ejb = rk eaj ejb ; (5.113)
T abk = eaiT ijk ejb = eairk ei b : (5.114)
These relationships enable the dependence between the innitesimal rotation
dab = d ba of the vector eai at innitesimal translations dsa to be established.
In fact, by (5.113) and (5.114), we have
dab = T abk dxk = Deaj ejb ; (5.115)
dab = T abk dxk = eaiDei b : (5.116)
where D is the absolute dierential [29] with respect to the Christoel symbols
i . Using (5.115), we can form the invariant quadratic form d 2 = da db
jk b a
to arrive at the Killing-Cartan metric
d 2 = dab dba = T abk T bandxk dxn = Deai Dei a (5.117)
5.5. A4 GEOMETRY . . . 29

with the metric tensor

Hkn = T abk T ban : (5.118)
Unlike metric (5.112), the metric (5.117) is specied on the rotations group
O (3:1) that acts on the manifold of the "rotational coordinates" eai .
Let us now introduce the covariant derivative

rm = rm + Tm ; (5.119)
where Tm is the matrix T abm with discarded matrix indices. We will regard
the components of the derivative as generators of the rotations group O (3:1).
Applying this operator to the tetrad ei that forms the manifold of "angular
coordinates" of the A4 geometry, we will arrive at

rm ei = rm ei + Tm ei = 0; (5.120)
hence
Tm = ei rm ei : (5.121)
It is interesting to note that, just as in (5.109) we have dened six "angular
coordinates" ei a through the four translational coordinates xi , so in (5:121) we
can dene 24 "supercoordinates" T abm through the six coordinates eia .
It follows from (5.120) that
rm ei = Tm ei : (5.122)
Recall that in the relationships (5.120)-(5.122) we have dened through rm
the covariant derivative with respect to ijk . We will now take the covariant
derivative rk of the relationships (5.122)
rk rm ei = rk (Tm ei ) = (rk Tm ei + Tm rk ei ) =
= (rk Tm ei + Tm ei ei rk ei ):
Using (5.121), we will rewrite this expression as follows
rk rm ei = (rk Tm Tm Tk )ei:
Alternating this expression in the indices k and m gives

r k rm ei = 12 Rkm ei ;
[ ] (5.123)

where
Rkm = 2r[mTk] + [Tm; Tk ]: (5.124)
Introducing in equations (5.124) the matrix indices (the bre indices), we
will obtain the structural equation of the group O (3:1)
Rabkm = 2r[m T ajbjk] + 2T ac[m T cjbjk] : (B )
30 CHAPTER 5. GEOMETRY OF ABSOLUTE. . .

It is easily seen that the structural equations of the rotations group (B ) coincide
with the second of Cartan's structural equations (5.124) of the geometry A4 .
In this case the quantities T abk and Rabkm vary in the rotations group O (3:1)
following the law 0 0 0
T ab0 k = aa T abk bb0 + aa ab0 ;k ; (5.125)
and appear as the potentials of the gauge eld Rabkm of the rotations group
O (3:1). In the process, the gauge eld of the group O (3:1) obeys the formula
0 0
Rab0 km = aa Rabkm bb0 : (5.126)
Note that the structural functions of the rotations group of A4 geometry are
the components of the curvature tensor Rabkm . It can be shown that the struc-
tural functions Rabkm of the rotations group O (3:1) satisfy the Jacobi identity
r nRajbjkm
[ ] + Rcb[km T ajcjn] T cb[nRajcjkm] = 0; (D )
which, at it was shown in the previous section, are at the same time the second
Bianchi identities of the A4 space.
Let us introduce the dual Riemann tensor
 1
Rijkm = 2 "spkm Rijsp; (5.127)

where "spkm is the completely skew-symmetrical Levi-Chivita tensor. Then the

equations (D) can be written as
  
rn R a b kn + R c b kn T acn T cbn R a c kn = 0 (5.128)
or, if we drop the matrix indices, as
rn R kn + R kn Tn 
Tn R kn = 0: (5.129)

5.6 Structural equations of A4 geometry in the

form of expanded, completely geometrized
Einstein-Yang-Mills set of equations
Einstein believed that one of the main problems of the unied eld theory
was the geometrization of the energy-momentum tensor of matter on the right-
hand side of his equations. This problem can be solved if we use as the space of
events the geometry of absolute parallelism and the structural Cartan equations
for this geometry.
In fact, folding the equations (B ); written as
Rijkm + 2r[k Tjij jm] + 2Tsi[k Tjsj jm] = 0 (5.130)
in indices i and k, gives
Rjm = 2r[i Tjijjm] 2Tsi[iTjsj jm]: (5.131)
5.6. STRUCTURAL EQUATIONS. . . 31

If then we fold the equations (5.131) with the metric tensor g jm, we have
R = 2g jm (r[iTjij jm] + 2Tsi[iTjsj jm]): (5.132)
Forming, using (5.131) and (5.132), the Einstein tensor
1
Gjm = Rjm g R;
2 jm
we obtain the equations
1
Rjm g R = Tjm ; (5.133)
2 jm
which are similar to Einstein's equations, but with the geometrized right-hand
side dened as

Tjm = 2 f(r[iT ijj jm] + T is[iT sjj jm])


1
g g pn (r[iT ijpjn] + T is[iT sjpjn])g (5.134)
2 jm
Using the notation
Pjm = (r[iT ijj jm] + T is[iT sjj jm])
then, by (5.134), we have
2 1
Tjm = (Pjm g g pn Ppn ): (5.135)
 2 jm
Tensor (5.135) has parts that are both symmetrical and skew-symmetrical
in indices j and m, i.e.,
Tjm = T(jm) + T[jm]: (5.136)
The left-hand side of the equations (5.133) is always symmetrical in indices
j and m, therefore these equations can be written as
1
Rjm g R = T(jm) ; (5.137)
2 jm
1
T[jm] =  ( ri
::i
jm rm Aj As
::s
jm ) = 0; (5.138)
where
Aj = Tjii : (5.139)
Relationship (5.138) can be taken to be the equations obeyed by the torsion
elds
::i
jm , which form the energy-momentum tensor (5.135).
i is skew-symmetrical in all the three indices,
In the case where the eld Tjk
we get
Tijk = Tjik = Tjki =
ijk : (5.140)
32 CHAPTER 5. GEOMETRY OF ABSOLUTE. . .

For such elds the equations (5.138) become simple, namely

ri
::i
jm = 0: (5.141)
The energy-momentum tensor (5.135) is symmetrical in indices j; m and
appears to be given by

Tjm = 1 (
::i 1

::s g
:ji
::s ): (5.142)
 sm ji 2 jm s ji
By (5.137), we have

Tjm = 1 (Rjm 1
g R): (5.143)
 2 jm
Using (5.131), (5.140) and (5.142) gives
Rjm =
::i ::s
sm
ji ; (5.144)
R = g jm
::i ::s :ji ::s
sm
ji =
s
ji : (5.145)
Substituting (5.144) and (5.145) into (5.143), we arrive at the energy-momentum
tensor (5.142).
Through the eld (5.140) we can dene the pseudo-vector hm as follows

ijk = "ijkm hm ;
ijk = "ijkm hm ; (5.146)
where "ijkm is the fully skew-symmetrical Levi-Chivita symbol.
In terms of the pseudo-vector hm we can write the tensor (5.142) as follows

Tjm = 1 (hj hm
1
g hi h ): (5.147)
 2 jm i
Substituting the relationships (5.146) into (5.141), we get
hm;j hj;m = 0: (5.148)
These equations have two solutions: the trivial one, where hm = 0, and
hm = ;m ; (5.149)
where is a pseudo-scalar.
Writing the energy-momentum tensor (5.148) in terms of this pseudo-scalar,
we will have
Tjm = 1 ( ;j ;m 1 gjm ;i ;i): (5.150)
 2
Tensor (5.150) is the energy-momentum tensor of a pseudo-scalar eld.
Let us now decompose the Riemann tensor Rijkm into irreducible parts
1
Rijkm = Cijkm + gi[kRm]j + gj [k Rm]i + Rgi[mgk]j ; (5.151)
3
5.6. STRUCTURAL EQUATIONS. . . 33

where Cijkm is the Weyl tensor; the second and third terms are the traceless
part of the Ricci tensor Rjm and R is its trace.
Using the equations (5.133), written as
 
1
Rjm =  Tjm g T ; (5.152)
2 jm
we will rewrite the relationship (5.151) as
1
Rijkm = Cijkm + 2g[k(iTj )m] T gi[mgk]j ; (5.153)
3
where T is the tensor trace (5.135).
Now we introduce the tensor current
1
Jijkm = 2g[k(iTj )m] Tg g (5.154)
3 i[m k]j
and represent the tensor (5.153) as the sum
Rijkm = Cijkm + Jijkm : (5.155)
Substituting this relationship into the equations (5.130), we will arrive at
Cijkm + 2r[k Tjijjm] + 2Tis[k Tjsj jm] = Jijkm : (5.156)
Equations (5.156) are the Yang-Mills equations with a geometrized source,
which is dened by the relationship (5.154). In equations (5.156) for the Yang-
Mills eld we have the Weyl tensor Cijkm , and the potentials of the Yang-Mills
eld are the Ricci rotation coeÆcients Tjk i .
We now substitute the relationship (5.155) into the second Bianchi identities
(D )
r[nRjij jkm] + Rsj[km Tjisjn] Tjs[nRjisjkm] = 0: (5.157)
We thus arrive at the equations of motion
r nCjij jkm + Cjs km Tjisjn Tjs nCjisjkm
[ ] [ ] [ ] = Jnijkm (5.158)
for the Yang-Mills eld Cijkm , such that the source Jnijkm in them is given in
terms of the current (5.154) as follows:
Jnijkm = r[nJjij jkm] + Jjs[km Tjisjn] Tjs[nRjisjkm] : (5.159)
Using the geometrized Einstein equations (5.133) and the Yang-Mills equa-
tions (5.156), we can represent the structural Cartan equations (A) and (B ) as
an extended set of Einstein-Yang-Mills equations

[ ]r keaj + T ikj eai = 0;

[ ] (A)
Rjm 12 gjm R = Tjm ; (B:1) (5.160)
i + 2r T i
Cjkm i s J ijkm ; (B:2)
[k jj jm] + 2Ts[k Tjj jm] =
34 CHAPTER 5. GEOMETRY OF ABSOLUTE. . .

in which the geometrized sources Tjm and Jijkm are given by (5.135) and (5.154).
For the case of Einstein's vacuum the equations (5.160) are much simpler

r k eaj + T ikj eai = 0;

[ ] [ ] (i)
Rjm = 0; (ii) (5.161)
C ijkm + 2r[k Tjijjm] + 2Tsi[k Tjsjjm] = 0: (iii)

The equations of motion (5.158) for the Yang-Mills eld Cijkm will then
become
r[nCjij jkm] + Cjs[km Tjisjn] Tjs[nCjisjkm] = 0: (5.162)
Equations (A) and (B:2) can be written in matrix form
r k eam
[ ] eb[k T ajbjm] = 0; (A)

C abkm + 2r[k T ajbjm] + 2T af [k T fjbjm] = J abkm ; (B:2)

where the current
1 a
J abkm = 2g[k (a Tb)m] T g [m gk]b ; (5.163)
3
is given by
1 1
T am =  (Ram 2 g am R); (B:1)
m = 0; 1; 2; 3; a = 0; 1; 2; 3:
By writing the equations (5.158) in matrix form, we have
r nC ajbjkm
[ ] + C cb[km T ajcjn] T cb[nC ajajkm] = J anbkm ; (5.164)
where
J anbkm = r[nJ ajbjkm] + J cb[km T ajcjn] T cb[nJ ajcjkm] : (5.165)
Dropping the matrix indices in the matrix equations, we have
r k em
[ ] e[k Tm] = 0; (A)
Ckm + 2r[k Tm] [Tk ; Tm] = Jkm ; (B:2)
rn C kn + [C kn ; Tn] =  J k ; (D )
 
where the dual matrices C kn and J k are given by

C kn = "knij Cij ;

J nk = "nkim Jim; (5.166)
  
J k = frn J kn + [J kn ; Tn]g: (5.167)
5.7. EQUATIONS OF GEODESICS. . . 35

For the Einstein vacuum we have


Rijkm = Cijkm =R ijkm = Cijkm ; (5.168)
therefore the equations (B:2) and (D ) become simpler
Ckm + 2r[k Tm] [Tk ; Tm] = 0; (B:2)
rnC kn + [C kn ; Tn] = 0: (D )
Using the formalismof external dierential forms, we can write the structural
equations (A) and (B:2) as follows:
dea eb ^ T ab = 0; (A)
C ab + dT ab T ac ^ T cb = J ab ; (B:2)
and the equations (D) as
dC ab + C af ^ T fb T fb ^ C af = N ab ; (D )
where
N ab = dJ ab + J af ^ T fb T fb ^ J af : (5.169)
Thus, the structural equations of A4 geometry, written as (5.160), represent
an extended set of Einstein-Yang-Mills equations with the gauge translations
group T4 dened on the base xi with the structural equations (A); and with
the gauge rotations group O (3:1), dened in the bre ei a with the structural
equations in the form of the geometrized equations (B:1) and (B:2):
5.7 Equations of geodesics of A4 spaces
The equations of geodesics for the geometry of absolute parallelism can be
obtained from the conditions of parallel vector displacement
dxi
ui = ds (5.170)

with respect to the connection of A4 geometry

ijk = ij + Tjk
i = ei ea :
a j;k (5.171)
In fact, we specialize the tetrad ei a so that the vector ei 0 would coincide
with the tangent to the world line, i.e.,
dxi
ei 0 = ui = ds : (5.172)

From the relationships (5:27) for the vector (5.172) we have


rk ui = ui;k +
ijk uj = 0 (5.173)
36 CHAPTER 5. GEOMETRY OF ABSOLUTE. . .

or
@ui i j i j
@xk + jk u + Tjku = 0: (5.174)
Multiplying this by uk = dxk =ds gives
dui + i uj uk i uj uk = 0
jk + Tjk (5.175)
ds
or, by (5.170),
d2 xi j k j k
i dx dx + T i dx dx = 0:
+ jk jk (5.176)
ds2 ds ds ds ds
These four equations (i = 0; 1; 2; 3) are the equations of geodesics of A4 space.
They are also the equations of motion for the origin O of tetrad ei a. Since in
the equations (5.176) the Ricci rotation coeÆcients Tjki have both symmetrical
and skew-symmetrical parts in indices j and k
Tjki = T(ijk) + T[ijk] =
=
::i im ::s
jk + g (gjs
mk + gks
mj );
::s (5.177)
T(ijk) = g im(gjs
::s ::s
mk + gks
mj ); (5.178)
T[ijk] =
::i
jk ; (5.179)
we can write the equations (5.176) as
d2 xi i dx
j dxk dxj dxk
+ T(ijk)
ds2 + ds ds = 0:
jk ds (5.180)
ds
Considering the structure of the equality (5.178), we will write it in the form
T(ijk) = g im(gjs
::s ::s im
mk + gks
mj ) = 2g
m(jk) ; (5.181)
hence the equations of geodesics for A4 space can be represented as
d2 xi + i dx
j dxk + 2g im
dxj dxk = 0:
jk ds m (jk) (5.182)
ds2 ds ds ds
For the terns in (5.181) we can introduce the following notation:

:ik:j = g imgks
::s i im ::s
jm ;
:jk = g gks
mj ;
i for space A will become
then the contorsion tensor Tjk 4

Tjki =
::i
jk
:ik:j +
i:jk ; (5.183)
where

:ik:j =
i:jk ;
whence
Tjki =
::i i
jk + 2
:jk : (5.184)
5.7. EQUATIONS OF GEODESICS. . . 37

The covariant dierential of an arbitrary vector v i with respect to the con-

nection (5.171) for parallel displacement from point xi to point xi + dxi becomes
Æv i = dv i +
ijk dxj = 0: (5.185)
If at an arbitrary point xi of A4 space we have two linear elements Æxi and
dxi and make a parallel translation of Æxi along the element dxi , then for the
nal point we will have [30]
xi + dxi + Æxi
ijkÆxk dxj = xi + dxi + Æxi + dÆxi : (5.186)
On the other hand, parallel translation of the vector dxi along the vector
Æxi gives
xi + Æxi + dxi
ijkdxk Æxj = xi + Æxi + dxi + Ædxi : (5.187)
Subtracting from the relationships (5.186) the equality (5.187), we get
dÆxi Ædxi = (
ijk Æxk dxj +
ijkdxk Æxj ) =
(
ijk
ikj )Æxk dxj = 2
i[jk]Æxk dxj =
= 2
::s k j ::s j k
jk Æx dx = 2
jk Æx dx : (5.188)
Let us now consider the variation of the integral
Z b
L(xi ; ui )ds; (5.189)
a
where ui is given by the relationship (5.170). We will write (5.188) as
Ædxi = dÆxi + 2
::s j k
jk Æx dx : (5.190)
Then at each point of the extremum we have
dxi d j dx
k
Æui = Æ ds = ds Æxi + 2
::i
jk Æx ds : (5.191)
Applying a common variational procedure to the integral (5.189), we get
Z b
ÆL(xi; ui )ds =
a
Z b


L(xi + Æxi ; ui + Æui) L(xi; ui ) ds =
a
Z b 
@L i @L i
=
@x i Æx + @ui
Æu ds = 0: (5.192)
a
Substituting here the relationship (5.191) gives
Z b 
@L @L d i @L ::i j k
@xi +
a @xi @ui ds @x + @ui 2
jk @x u ds = 0:
38 CHAPTER 5. GEOMETRY OF ABSOLUTE. . .

We now integrate by parts the second term here to obtain

Z b 
@L d @L @L k
+ 2
::j i
a @xi ds @u i ik @uj u @x = 0
or, since @xi is arbitrary, we arrive at [30]
d @L @L @L k
+ 2
::j
@uj u = 0:
ki (5.193)
ds @ui @x i

Let now
L = (gik ui uk )1=2 ; (5.194)
along the extremum L = 1 by the relationship
gik ui uk = ui ui = 1:
Substituting the Lagrangian (5.194) into equations (5.193) gives
i
gmi du + mjk u
j uk + 2
::s k j
mj gsk u u = 0: (5.195)
ds
Multiplying this relationship by g im, we get
dui + i uj uk
kj + 2g imgks
::s j k
mj u u = 0
ds
or
dui i uj uk
+ kj + 2g im
m(jk) uj uk = 0: (5.196)
ds
We have thus obtained, using the variational principle, the equations of
the geodesics in the form (5.182). Consider now the equations that describe
the variation of the orientation of the tetrad ei a as it moves according to the
equations of the geodesics (5.196). We will rewrite the equations (5.43) as
@k ei a +
ijk eja = 0
or
dei a +
ijkeja dxk = 0: (5.197)
Dividing these equations by ds yields
dei a +
i ej dxk = 0: (5.198)
ds jk a ds

Further, taking the second derivative d2 ei a=ds2 , we will have

   
d dei a d @ei a dxk @ 2 ei a dxk dxm @ei a d2 xk
@xm @xk ds ds + @xk ds2 :
= = (5.199)
ds ds ds @xk ds
5.7. EQUATIONS OF GEODESICS. . . 39

Since
@ 2 eia @ i i j
@xm @xk = @xm (
ak ) =
jk;m e a

isk (
sjmeja ) = (
ijk;m +
isk
sjm)eja
and
@eia d2 xk =
i
s dxk dxm ej ;
@xk ds2 js km ds ds a
we have
d2 ei a i i
s
k m
i
s ) dx dx ej = 0:
+ (
jk;m
sk jm
js km ds ds a (5.200)
ds2
Substituting here the sum (5.171), we have
d2 ei a i i i s i Ts
+( jk;m + Tjk;m sk jm sk jm
ds2
Tski s
jm Tski Tjm
s i
js
s
km Tjsi skm
i Ts dxk dxm j
js km Tjsi Tkm
s ) e = 0: (5.201)
ds ds a
Since independent equations (5.201) (for three Euler's angles and three
pseudo-Euclidean angles) describe the variation of the orientation of tetrad ei a
as it moves from the origin O according to the equations of geodesics (5.196).
In A4 spaces, where the metric is at
gik = ik = diag(1 1 1 1); (5.202)
the Christoel symbols i js vanish and the equations (5.201) become
d2 ei a + (T i k m
s ) dx dx ej = 0;
jk;m Tski Tjm
s Tjsi Tkm (5.203)
ds 2 ds ds a
and the equations of geodesics (5.175) will become
d2 xi j k
i dx dx = 0:
+ Tjk (5.204)
ds 2 ds ds
We now introduce the tensor of the four-dimensional angular velocity of
rotation tetrads ei a [33]
dxk deia a deja a

ij = Tijk
ds =
ds e j = ds e i (5.205)
with the symmetry properties

ij =
ji ; (5.206)
determined by the symmetry (5.48), for which the Ricci rotation coeÆcients
hold.
40 CHAPTER 5. GEOMETRY OF ABSOLUTE. . .

Using (5.205), we will write the equations (5.204) and (5.203) as

d2 xi +
i dxj = 0;
j ds (5.207)
ds2
d
i j dxk dxm dx
T ijk;m ds ds + T ijs
sk dsk = 0: (5.208)
ds
The skew-symmetric matrix (5.206) can be represented as
0 1
0
01
02
03
B
10 0
12
13 C

ij = B
@
20
21 0
23
C
A (5.209)

30
31
32 0
Let us now give a physical interpretation of the components of the matrix
(5.209). We multiply the equations (5.207) by the mass m and rewrite them as
j
m d x2i + m
ij
2
dx = 0: (5.210)
ds ds
If the condition (5.202) holds, there equations can be represented as
j
m du i + m
ij dx = 0; (5.211)
dso dso
where
dso = (ik dxi dxk )1=2 (5.212)
is the pseudo-Euclidean metric and ui = dxi =dso .
We represent the equations (5.211) in the form
du i dxj dxk
m = mTi(jk) ; (5.213)
dso dso dso
where the part of T symmetric in indices j and k is given by (5.178).
Assuming that motion governed by the equations (5.213) is nonrelativistic
(v=c  1), we will write the three-dimensional part of these equations as
du  dxo dxk dx dxk
m = mT(ok ) 2mT(k) (5.214)
dso dso dso dso dso
or, from the relationship (5.205), as
o dx :
m du  = m
o dx 2m
 (5.215)
dso dso dso
Since in the nonrelativistic approximation
v
dso = cdt; ua = ;
c
5.7. EQUATIONS OF GEODESICS. . . 41

and dxo = cdt, the equations (5.215) can be written as

dv 1 dx
m = mc2
o 2mc2
 : (5.216)
dt c dt
It is known from classical mechanics that the nonrelativistic equations of
motion of the origin O of a three-dimensional accelerated reference frame under
inertia forces alone have the form [34]
d
dt (mv) = m ( W + 2[v! ]) ; (5.217)

where W is the vector of translational acceleration, and ! is the vector of the

three-dimensional angular velocity of rotation of the accelerated reference frame.
We write these equations as
 
d (mv ) = m Wo + 2! dx ; (5.218)
dt  dt
where W = (W10; W20 ; W30);
0 1
0 !3 !2
! = ! = @ !3 0 !1 A (5.219)
!2 !1 0

! = (!1 ; !2 ; !3 );
and comparing these with (5.217), we obtain
W1 ;
= W2 ;
= W3 ;

10 =
c2 20
c2 30
c2
! 3 ;
= !2 ;
= ! 1 :

12 =
c 13
c 23
c
Therefore, the matrix (5.209) in this case has the form
0 1
0 W1 W2 W3
1 B W1 0 c!3 c!2 C

ij = B
W2 c!3 0 c!1
C (5.220)
c2 @ A
W3 c!2 c!1 0

It is seen from this matrix that the four-dimensional rotation of the tetrad
ei a, caused by the torsion of the A4 spaces, gives rise in physics to inertia elds
associated with translational and rotational accelerations.
42 CHAPTER 5. GEOMETRY OF ABSOLUTE. . .

5.8 Structural equations of right and left A4 geo-

metry
We can consider three forms of the geometry of absolute parallelism.
(1) A4 geometry, with the nonzero Riemannian tensor Ri jkm and torsion
::i

jk . The structural Cartan equations then become

r k eaj
[ ] + T i[kj ]eai = 0; (5.221)
Ri jkm + 2r[k Tjij jm] + 2Tsi[k Tjsj jm] = 0: (5.222)
(2) A4 geometry, with the zero Riemannian Ri jkm and nonzero torsion
::i
jk .
In that case the structural Cartan equations can be written as
r k eaj + T ikj eai = 0;
[ ] [ ] (5.223)
r k Tjij jm + Tsi kTjsj jm
[ ] [ ] = 0: (5.224)
(3) A4 geometry, with the zero Riemannian tensor Ri jkm and noncoordinate
torsion
::i
jk . The structural Cartan equations of the geometry coincide with the
structural equations of the pseudo-Euclidean space E4, and they look like
Æ Æ
r[k eÆ aj]+ T i[kj ] eÆ ai = 0; (5.225)
Æ Æ Æ Æ
r k T ijj jm + T is k T sjj jm
[ ] [ ] = 0; (5.226)
Æ
where the tetrad e ai determines the "coordinate torsion"
Æ Æi Æa 1 Æi Æa Æa

::i
jk =e a e [k;j ] = 2 e a (e k;j e j;k ): (5.227)

Since in the pseudo-Euclidean space the T4 and O (3:1) groups hold globally
and its internal geometry is trivial, then, for example, in the Cartesian coor-
dinate x0 = ct, x1 = x, x2 = y , x3 = z the structural equations (5.225) and
(5.226) become the identities
0  0; (5.228)
0  0: (5.229)
If we now go over to the spherical coordinates
x0 = ct; x1 = r; x2 = ; x3 = ';
we will get the equations (5.225){(5.227), which include:
(a) components of the \coordinate" tetrad
Æe (0) eÆ (1) Æ Æ
0 = 1 = 1; e (2) 2 = r; e (3)
3 = r sin ;
Æe 0 eÆ 1 Æe 2 1 Æ3 1
(0) = (1) = 1; (2) = ; e (3) = ; (5.230)
r r sin 
5.8. STRUCTURAL EQUATIONS OF THE RIGHT. . . 43

(b) components of \coordinate torsion"

Æ Æ Æ 1

::212 =
::313 = (2r ) 1 ;
::323 = 2 cot  ; (5.231)

(c) components of the Ricci rotation coeÆcients

Æ Æ Æ
T 122 = r; T 133 = r sin2 ; T 233 = sin  cos ;
Æ Æ 1 Æ
T 212 =T 313 = ; T 323 = cot :
r (5.232)

Using the formulas

gÆ ik = ab eÆ ai eÆ bk ; ab =  ab = diag(1 1 1 1);
we nd the components of the metric tensor
gÆ 00 =gÆ 11 = 1; gÆ 22 = r; gÆ 33 = r 2 sin2 ;
the metric
Æ
ds2o =gij dxi dxj = c2 dt2 dr 2 r 2(de2 + sin2 d'2 )
and the components of the Christoel symbols
Æ1 Æ1 Æ2 Æ3
33 = r sin2 ; 22 = r; = 13 = 1r ; (5.233)
Æ2 Æ 3 12
33 = sin  cos ; 23 = cot :

Thus, in the pseudo-Euclidean geometry A4, when we deviate from Cartesian

coordinates, instead of the identities (5.228) and (5.229) we get the "coordinate
structural equations" (5.225) and (5.226).
Suppose now that the initial pseudo-Euclidean space A4 is deformed in a
continuous manner (e.g., using conformal transformations) into an A4 space
with a nonzero dynamic torsion eld and the structural equations (5.223) and
(5.224). We can distinguish the right
1 i a

::i i a r aj;k )
+
jk = r ar [k;j ] = 2 r a(r k;j (5.234)

and left
1i a

::i i a a
jk = l a l [k;j ] = 2 l a (l k;j l j;k ) (5.235)
torsion elds. In these equations rai and lia stand for the right and left tetrads.
respectively.
We well take the right tetrad rai to mean a tetrad e i a, such that when the
+

three-dimensional spatial part rotates from the x axis to the y axis the vector
of the angular rotational velocity points along the z axis, so that the rotation
occurs counterclockwise if looking from the side to which the z vector points.
44 CHAPTER 5. GEOMETRY OF ABSOLUTE. . .

For example, the four-dimensional rotation matrix (5.220) for the right tetrad
looks like 0 1
0 W1 W2 W3
+ 1 B W1 0 c!3 c!2 C

ij = c2 B
@ W2 c!3 0 c!1 A
C (5.236)
W3 c!2 c!1 0
whereas for the left rotations we have
0 1
0 W1 W2 W3
1 B W1 0 c!3 c!2 C
C:

ij = B
W2 c!3 0 c!1 (5.237)
c2 @ A
W3 c!2 c!1 0
It is seen that +

ij =
ij : (5.238)
From (5.205) and (5.238), we have
i T ijk :
+
T jk = (5.239)
Since the metric tensor gik is determined both by the right and left tetrad
in a similar manner [35]
gik = ab riarkb = ab lai lbk ; (5.240)
it follows from the denition
Tjki =
::i im ::s ::s
jk + g (gjs
mk + gks
mj ) (5.241)
that the components (5.234) and (5.235) of the right and left torsion elds dier
in sign

::i ::i
+
jk =
jk : (5.242)
By dividing the torsion elds into left- and right-hand ones, we thereby split
+
the translations group T4 into the right T 4 and left T 4 translations groups; and
the rotations group O (3:1) into the right SO + (3:1) and left SO (3:1) rotations
group.
We will write the structural Cartan equations of the A4 geometry, which are
transformed using continuous transformations in T4 and SO + (3:1) groups, as
follows:
r[k +e aj]+ T i[kj ] +e ai = 0;
+
(5.243)

r k T ijj jm + T is k T sjj jm
+ + +
[ ] [ ] = 0: (5.244)
Accordingly, the equations
r ke aj + T ikj eai = 0;
[ ] [ ] (5.245)
5.8. STRUCTURAL EQUATIONS OF THE RIGHT. . . 45

r kT ijj jm + T is k T sjj jm
[ ] [ ] =0 (5.246)
are transformed continuously in the T4 and SO + (3:1) groups.
It is clear that discrete transformations | inversion transformations | en-
able us to transform the right equations (5.243) and (5.244) into left equations
(5.245) and (5.246), and vice versa.
The property (5.242) of the A4 geometry enables an empty pseudo-Euclidean
geometry to be "split" into right- and left-hand geometries:
Æ ::i

::i ::i
+
jk =
jk +
jk = 0; (5.247)
whose torsion is nonzero. This property appeared to be quite useful for the
description of the production of matter from "nothing" in the theory of physical
vacuum [36].
If now we split the structural Cartan equations (5.221) and (5.222) into right
and left ones, we will get

r k e aj + T ikj e ai = 0;
+ + +
[ ] [ ] (5.248)

i s
R i jkm + 2r[k T ijj jm] + 2 T s[k T jj jm]= 0;
+ + + +
(5.249)

r k e aj + T ikj e ai = 0;
[ ] [ ] (5.250)

R i jkm + 2r[k T ijj jm] + 2 T (5.251)i

s[k T sjj jm] = 0:
Writing the structural Cartan equations as the extended right and left Einstein-
Yang-Mills equations, we will arrive at

r k e aj + T ikj e ai = 0;
+ + + +
[ ] [ ] (A)
+ +
R jm 2 gjm R=  T jm;
1
+ +
(B :1) (5.252)
C ijkm + 2r[k T ijj jm] + 2 T is[k T sjj jm] =  J ijkm : (B :2)
+ + + + + +

r k e aj + T ikj e ai = 0;
[ ] [ ]
(A)
R jm gjm R=  T jm;
1
(B :1) (5.253)
2

C jkm + 2r[k T ijj jm] + 2 T is[k T sjj jm] =  J

i i
jkm : (B :2)
In the theory of physical vacuum that is based on the universal relativity
principle [37], equations (5.252) and (5.253) describe the right and left matter
produced from vacuum.
46 CHAPTER 5. GEOMETRY OF ABSOLUTE. . .
Chapter 6

The geometry of absolute

parallelism in spinor basis
6.1 Three main spinor bases of A4 geometry
The geometry of absolute parallelism, as laid down in vector basis, enables the
structural equations of this geometry of be represented as right (invariant with
respect to the T4+ and SO + (3:1)) groups and left (invariant with respect to the
T4 and the SO (3:1) groups) groups of the structural equations (A+ ), (B + )
and (A ) and (B ), respectively. Equations (A+), (B + ) (or (A ), (B )) can,
in turn, be split by a transition into a group of equations, whose component
elds have opposite spins. For this purpose, we have to use spinor basis and
some elements of spinor analysis.
We will view the spinor geometry A4 as a dierentiable manifold X4 , such
that at each point M with the translational coordinates x (i = 0; 1; 2; 3) a two-
dimensional spinor space C 2 is introduced [38]. There are three possibilities for
introducing the spinor basis in the spinor space C 2 :
(a) spinor -basis formed by the Infeld-Van der Werden symbols i _ [39],
which satisfy the equality
rn i _ = 0; (6.1)

(b) spinor
-basis formed by the Newman-Penrose symbols Ai B_ [40], which
satisfy the equality

rn Ai B_ = 0; (6.2)

(c) spinor dyad basis B , which satises the equality [41]

"BD D rk B = 0: (6.3)
_ : : : and A; B;
In relationships (6.1){(6.3) the indices ; ; _ : : : are spinor in-
_ _
dices that take on the values 0,1 and 0; 1. Any local vector Ai that belongs to

47
48 CHAPTER 6. GEOMETRY OF ABSOLUTE. . .

C 2
can be represented as a spin-tensor of the second rank either in the spinor
- basis
Ai = A_ i _ ; (6.4)
or in the spinor
-basis
Ai = AAB_ Ai B_ : (6.5)
All the spin-tensors associated with the -basis will have the spinor indices
; ;_ : : :, and the spin-tensors associated with
basis will have spinor indices
A; B; _ : : :. As to dyad B , it is a connection between - and
-basis

Ai B_ = i _ A B :
_
(6.6)
Here
 B = B ;
_

and the bar on the right-hand side of the equality implies complex conjugation.
Spinor
-basis is connected with the vector basis eai by
Ai B_ = eia Aa B_ ; (6.7)

iAB_ = eai aAB_ ; (6.8)

where i AB_ are complex Hermitian (i AB_ = i AB
_
) matrices, and the matrices
a A B
 AB_ and a have the form
_

0 1
1 0 0 1
= B 0 1 1 0 C
 aAB_ = (2) 1 2 B
@ 0 i i 0
C;
A (6.9)
1 0 0 1
0 1
1 0 0 1
aAB_ = (2) 1 2= B
B 0 1 i 0 C
C; (6.10)
@ 0 1 i 0 A
1 0 0 1
where
det(Aa B_ ) = i; det(aAB ) = i:
_

From the orthogonality conditions for the tetrad eia

eaieja = Æi j ; eaiei b = Æ ab (6.11)
and the relationships (6.7){(6.10) follows the orthogonality conditions for the
spinor
-basis
iAB Aj B_ = Æi j ;
_
(6.12)
iAB Ci E_ = Æ AC Æ B_ E_ :
_
(6.13)
6.1. THREE MAIN SPINOR. . . 49

For the spinor -basis the following orthogonality conditions hold [54]
i_ j _ = Æi j ; (6.14)

i_ i _ = Æ   Æ _ _ : (6.15)
Whence, by (6.6) and (6.12)-(6.13), follow the orthogonality conditions for
the spinor dyad
o 1 = 1;
o  = o o = 0;
 o (6.16)
1 1 = 0:
In addition, there are the relationships [54]
o o 1 o = Æ ; (6.17)
o 1 1 o = " ;
where  
0 1
" = "= " _ Æ_ = = " _ Æ_
1 0 (6.18)
is the fundamental spinor [40] that obeys the following relationships:
" " = " = " ; (6.19)

" " = Æ Æ Æ Æ ; (6.20)

" = 2; (6.21)

"[ "Æ ] = 0; (6.22)

 
1 0
" = 0 1 : (6.23)
The fundamental spinor " increases and decreases the indices on the spin-
tensor associated with the -basis, similar to the metric tensor gik in the vector
basis. In the spinor
-basis it has the form
"AB = " A B ; (6.24)
so that  
0 1
"AB = "AB = "C_ D_ = "C_ D_ = 1 0 : (6.25)
The fundamental spinor "AB increases and decreases indices on the spin-
tensors associated with the
-basis. For example, we have
:::A::: :::::: AB ::: ::: :::A:::
::: ::: "AB = :::B:::; " :::B::: = :::::: ;
:::A::: :::::: A_ B_ ::: ::: (6.26)
':::::: "A_ B_ = ':::B:::
_
_ " ':::B::: _ = ':::A:::
:::::: :
_
50 CHAPTER 6. GEOMETRY OF ABSOLUTE. . .

If the spinor is skew-symmetric in two indices

:::A:::B::: = :::B:::A:::; (6.27)
then, using the fundamental spinor "AB , it can be represented as [40]
1 ::: :::C::: :
:::A:::B::: = "AB :::C::: (6.28)
2
The same properties are valid in the spinor -basis for the fundamental spinor
" .

6.2 Spinor representation of the structural Car-

tan equations of A4 geometry
The relationship (6.28) makes it possible to reduce spinors skew-symmetric
in primed and unprimed indices to spinors that are completely (or partially)
symmetrical in primed and anprimed indices. In the space of spinors of this
type irreducible representations of the groups SL(2:C ) are realized [42]. This
group replaces the group SO (3:1) on passing over to the spinor basis.
Denition 6.1. We will say that the components of a spinor with r symmet-
rical lower indices and with s symmetrical lower primed indices are transformed
in D (r=2;s=2) irreducible representation of the group SL(2:C ).
For example, the spinor
FAB = FBA
is transformed in D (1:0), and the spinor
FC_ D_ = FD_ C_
in the D (0:1) irreducible representation of the group SL(2:C ).
We will write the main relationships of the A4 geometry in the spinor
-
basis. This can be accomplished using the spinor representation of the arbitrary
:::i:::::: in the
-basis
tensor T::::::j:::
:::AB:::::: j
T::::::C = iAB T::::::i::::::
:::j:::C E_
_ _
E:::
_ (6.29)
or simply replacing the matrix indices by two spinor ones as follows:

eai $ iAB ;
_
(6.30)

T abm $ T AB_ C Dm
_ ; (6.31)
Ra bkm $ RAB C Dkm
_ ;_
(6.32)
ab $ ABC _ = "AC "B_ D
_ D _; (6.33)
and so on.
6.2. SPINOR REPRESENTATION. . . 51

Proposition 6.1. In the spinor
-basis the metric tensor gij of the A4
geometry has the form
gij = "AC "B_ D_ iAB_ jC D_ : (6.34)
Proof. Substituting into
gij = abeai ebj
the relationships (6.7) and (6.8) written as
eai = iAB_  a AB_ ; ebj =  C D_ j  b C D_ ; (6.35)
we have
gij = ab iAB_ Aa B_ jC D_ Cb D_ : (6.36)
From the relationships (6.9), (6.10), (6.25) and the denition
ab =  ab = diag(1 1 1 1);
we obtain the following equality:
abAa B_ Cb D_ = "AC "B_ D_ :
Substituting this into (6.36), we arrive at the formula (6.34).
We now write the structural Cartan equations in matrix form
r k eam ebk Tjabjm = 0;
[ ] [ ] (A)
Rabkm + 2r k Tjabjm + 2Tcak Tjcbjm
= 0:
[ ] [(B ) ]

Using the rules (6.30)-(6.32), we write these equations in the spinor
-basis
r k mAB  CkD T AjBC D jm
[ ]
_
[
_ _
_ ]
= 0; (6.37)

RAB_ C Dkm
_ + 2r[k T AB_ jC D_ jm] + 2T AB_ E F_ [k T E F_ jC D_ jm] = 0: (6.38)
Consequently, the second Bianchi identity of the A4 geometry
r nRajbjkm
[ ] + Rcb[km Tjacjn] Tbc[nRajcjkm] = 0 (D )
in the spinor
-basis becomes
r nRABjC D jkm
[
_
_ ]
+ RE FC_ D_ [km T AjBC_ D_ jn] T ECF_D_ [nRAEj_E F_ jkm] = 0: (6.39)

Proposition 6.2. If Fij = Fji is a real skew-symmetrical tensor, then

the corresponding spinor
FABC i j
_ D_ = Fij AB_ CD
_ (6.40)
can be represented in the form
1
FABC
_ D_ = ("B_ D_ FAC + "AC F B_ D_ ); (6.41)
2
52 CHAPTER 6. GEOMETRY OF ABSOLUTE. . .

where the spinor

FAC = FCA (6.42)
is transformed in the D (1:0) irreducible representation of the group SL(2:C ),
and the spinor
F BD = FB_ D_ = FD_ B_ (6.43)
in the D (0:1) irreducible representation of the same group.
Proof. Since the tensor Fij is skew-symmetric, we have, by (6.40),
FABC
_ D_ = FC DA
_ B_: (6.44)
We rewrite this as
1 1
FABC _ =
_ D (F _ D_ FC DA _ = (FABC FC DA
_ B_ + FC DA
2 ABC _ B _ D_ _ B_
2
FC DA
_ B _ ): (6.45)
Using the fundamental spinor (6.25), we can write (6.45) as follows:
1
FABC _ = (" F F + "B_ D_ FAEC E_ ): (6.46)
_ D
2 AC F B_ D_ _

Denoting FAC = (1=2)FAEC _

E_ , we have, by (6.44),

1 E_ = 1 F E_ _ = F :
FAC = 2 FAEC (6.47)
_
2 A CE CA

Further, introducing the notation F B_ D_ = 21 FF B_ F D_ and considering that Fij

is real, we nd
1 1
F B_ D_ = 2 FF B_ F D_ = 2 F F_ B F_ D = F BD : (6.48)
Substituting the relationships (6.47) and (6.48) into (6.46), we arrive at
(6.44). By denition, the spinor FAC = FCA belongs to the D (1:0) irreducible
representation of the groups SL(2:C ); and spinor F B_ D_ = F D_ B_ { to the D (0:1)
irreducible representation of the group.
Since the quantities TABC _ Dm_ and RABC _ Dkn
_ in the equations (6.37) and
(6.38) are skew-symmetric in the pair of spinor matrix indices AB_ and C D_ , we
can represent them, by (6.27){(6.28), as
1
TABC _ Dk = ("B_ D_ TACk + "AC TB+_Dk _ ); (6.49)
_
2
1
RABC _ Dkn_ = ("B_ D_ RACkn + "AC R+B_ Dkn ); (6.50)
2 _

where
1 1
TACk = 2 "B_ D_ TABC _ Dk _ ; TB+_Dk = "AC TABC _ Dk_ ; (6.51)
_
2
1 1
RACkn = "B_ D_ RABC _ Dkn ; R+B_ Dkn = "AC RABC _ Dkn : (6.52)
2 _ _
2 _

In these relationships the + sign with the spinor matrices implies Hermitian
conjugation.
6.3. SPLITTING OF THE STRUCTURAL. . . 53

6.3 Splitting of structural Cartan equations

into irreducible representations of the
group SL(2:C)
Matrices (6.51) and (6.52) can be transformed in the spinor indices as follows:
0 0 0
T AC 0 k = SAA T ACk S CC 0 + SAA S AC 0 ;k ; (6.53)
0 0 0
T +D_B0_k = SB+_ B_ T +Dk
B_ +D_ + S +B_ S B_ ;
_ SD_0 B_ D_ 0 ;k (6.54)
0 0
RA C 0 kn = SAA RA Ckn SCC0 ; (6.55)
0 0
R+B_ D_ 0 kn = SB+_ B_ R+B_ Dkn
_ SD+_ D0 _ : (6.56)
0
Matrices of the transformations SAA and SB+_ B form the group SL(2:C ), and the
matrices 0
SAA (6.57)
form the subgroup
SL+ (2:C ) (6.58)
of the group SL(2:C ), in which the spinors belonging to the irreducible repre-
sentation D (r=2; 0) are transformed.
On the other hand, the matries
0
S +B_ B_ (6.59)
form the subgroup
SL (2:C ) (6.60)
of the group SL(2:C ), in which the spinors belonging to the irreducible rep-
resentation D (0; s=2) are transformed. These properties of the spinors enable
the structural Cartan equations to be split into equations that contain spinors
transformed in D (r=2; 0) or D (0; s=2) irreducible representations of the group
SL(2:C ):
Proposition 6.3. The second structural Cartan equations (B ) in the
spinor
-basis are split into the equations of the form
RACkn + 2r[kTjAC jn] + 2TAE [k TjEC jn] = 0; (6.61)

R+B_ Dkn
_ + 2r[k Tj+B_ D_ jn] + 2TB+_F_ [k T +jFD__ jn] = 0: (6.62)
Proof. We write the second structural Cartan equations (6.38) as
BABC
_ Dkn
_ = RABC
_ Dkm
_ + 2r[k TjABC
_ D_ jm] + 2TABE
EF_
_ F_ [k TjC D
_ jm] = 0: (6.63)
54 CHAPTER 6. GEOMETRY OF ABSOLUTE. . .

Using the fact that the spinor BABC_ Dkn

_ is skew-symmetric in the pair of
spinor indices AB_ and CD
_ , we will write it in the form
1
BABC
_ Dkn
_ = ("B_ D_ BACkn + "AC BB+_Dkn ) = 0; (6.64)
2 _

where
1
BACkn = "B_ D_ BABC
_ Dkn
_ = 0; (6.65)
2
1
BB+_Dkn = "AC BABC _ Dkn = 0: (6.66)
_
2 _

Substituting (6.61) into the equations (6.65) and (6.66) and using the ma-
trices (6.51) and (6.52), we will arrive at the structural equations (6.61) and
(6.62) in split form. In the derivation we have used the properties (6.19)-(6.23)
of the fundamental spinor "AB .
Proposition 6.4. Matrices TACk and TB+_Dk _ in the dyad basis C have
the following form:
TACk = C rk A = TACk ; (6.67)
_
TB+_Dk
_ =  _ D_ rk  B_ = TB+_Dk
_ : (6.68)
Proof. We write the matrices
Tabk = ei b rk eai
in the spinor basis, using the rules (6.30) and (6.31)
TABC
_ Dk
_ = Ci D_ rk ABi
_ : (6.69)
Substituting this expression into the rst one of (6.51) gives
1
TACk = 2 "B_ D_ Ci D_ rk ABi_ : (6.70)

Using the formula (6.6), we write ABi

_ as

ABi_ = _ iA  B_ :
_
(6.71)
Substituting (6.71) into (6.70), we have
1 _ 1 _
TACk = "B_ D_ Ci D_ rk (i_ A B_ ) = "B_ D_ Ci D_ i_ rk (A B_ ); (6.72)
2 2
since rk (i_ ) = 0.
Further, considering that

Ci D_ i_ =  _ i C  D__ i_ =

_
= Æ Æ _ _ C  D_ = C  D_ _ ;
6.3. SPLITTING OF THE STRUCTURAL. . . 55

we will write (6.72) as

1
TACk = 2 "B_ D_ C  D_ _ ( B_ rk A + A rk  B_ ):
_ _
(6.73)

In the dyad basis we have the equalities

"B_ D_ =  D_ _ B_ ; "B_ D_  D_ _ rk B_ = 0;
_ _

which are conjugates of (6.3) and (6.24). Using these equalities, we can easily
obtain (6.67). Similarly, for the conjugate matrix T + B_ Dk
_ , we have (6.68).

Proposition 6.5. In the spinor
-basis the rst structural Cartan equa-
tions (A) of the A4 geometry have the form
r k Ci D T kjCE  E D ji jCi F j Tk D F = 0
[
]
_ [ _
] [
_
+ _
] _
(6.74)
or, dropping the matrix indices,
r k  i T k  i  iTk
[
]
[
] [ +
] = 0: (6.75)
Proof. Let us take the derivative rk Ci D_ :

rk Ci D = rk (i  C D ) = i  (D rk C + C rk D ):
_ _ _
_ _ _ _ _ _

Using (6.67) and (6.68), we will write this relationship as

rk Ci D = i  (TCEk  E  D + TDFk C  F ):

_ _ _
_ _ _
+
_ _ (6.76)
Here we have used the normalization conditions
E  E = 1;  _ F_  F = 1:
_ _

Multiplying the terms on the right-hand side (6.76) we obtain, from (6.71),
rk Ci D TCEk DiE CiF TDFk = 0
_ _
_ +
_ _ (6.77)
or
rk Ci D TkCE DEi Ci F TkDF :
_ _ (6.78) _
+_
_

Alternating this relationship in the indices k and i, we obtain the equations

(6.74).
Proposition 6.6. The second Bianchi (D ) identities of the A4 geometry
in the spinor
-basis are split into the following equations:
  
rn R ACkn R ECkn T E An + R EAkn T E C n = 0; (6.79)
  
rn R +B_ Dkn
_ R +F_ Dkn
_ TB+_ Fn + R +F_ Bkn
_
_ TD+_ F n = 0:
_
(6.80)
56 CHAPTER 6. GEOMETRY OF ABSOLUTE. . .

Proof. Increasing and decreasing, using the metric tensors ab and gik , the
tensor indices in the identities (6.150), we will write them in the form
  
rn Rabkn Rcbkn T ca n + Rackn T cb n = 0: (6.81)
In this equality we now pass over to the spinor indices using (6.31) and (6.32)
go get

rn R ABC  n  n
_ Dkn
_ RE FC
_ Dkn
_ T EF
_
AB_ + RE F_ ABkn
_ T EF
_
C D_ = 0: (6.82)
We now write this relationship in the form
DAn BC
_ Dkn
_ = 0; (6.83)
where by DAn BC_ Dkn
_ we have denoted all the terms on the left-hand side of (6.82).
Since the relationship (6.83) are skew-symmetrical in the pair of indices AB_ and
C D_ , we will write it in the form
1
DAn BC
_ Dkn
n
= ("B_ D_ DACkn + "AC D +n B_ Dkn ) = 0; (6.84)
_
2 _

where
n 1 1
DACkn = "B_ D_ DAn BC
_ Dkn = 0; DB+_ nDkn = "AC DAn BC
_ Dkn = 0:
2 _ _
2 _

Substituting here (6.82), we will get (6.79) and (6.80).

Physically, the spinor splitting of the structural Cartan equations (A) and
(B ) implies splitting into the equations of \matter " and \skew-symmetry", just
as it has been done by Dirac in his derivation of equations for the electron and
the positron. We can now write equations that are transformed in the groups
SL+ (2:C ) as
r[kCi]D_ T[kjCE DE_ ji]  [i jC F_ j Tk+]D_ F_ = 0; (As)
RACkn + 2r[k TjAC jn] + 2TAE [k TjEC jn] = 0; (B s+ )
and in the group SL (2:C ) as

r k  i C D T kjCE DE ji  i jC F j Tk D F
[
]
_ [ _
] [
_
+ _
] _
= 0; (As)

R+B_ Dkn + 2r[k Tj+B_ D_ jn] + 2TB+_F_ [k Tj+D_Fjn] = 0: (B s )

_
_

In the numerations of these formulas s implies transformation in a spinor

group. Dropping the matrix indices, we will write these relationships as
r k i
[
]
T[k  i]  [iTk+] = 0; (As)

Rkn + 2r[kTn] [Tk ; Tn] = 0; (B s+ )

6.4. CARMELI MATRICES. . . 57

r k  i T k  i  iTk = 0;
[
]
[
] [ +
] (As)
Rkn + 2r k Tn [Tk ; Tn ] = 0:
+
[
+
]
+ +
(B s )
Correspondingly, discarding the matrix indices in the equations (6.79) and
(6.80), we obtain
 
rn Rkn +[R kn ; T n] = 0; (D s+)
 
rn R +kn + [R +kn ; T +n ] = 0: (D s )
6.4 Carmeli matrices
Equalities (6.67) and (6.68) can be written in matrix form
Tk =  rk ; (6.85)
Tk+ =  + rk  + ; (6.86)
where Tk and  are 2  2 complex matrices with elements T A Bk and Aa , respec-
tively. Multiplying Tk by  k AB_ , we can introduce the traceless Carmeli 2  2
matrices [44-46]
TAB_ = Ak B_ Tk ; (6.87)
A; C : : : = 0; 1; B;_ D_ : : : = 0_ ; 1_
with the components
   
T00_ = "  ; T01_ =   ;
  "     
   (6.88)
T10_ =   ; T11_ =  :

Using matrices (6.87), we can dene the matrix elements

CD
AB_ 00 01 10 11
00_ "   "
(TAB_ )C D = 01_     ; (6.89)
10_    
11_  
where (TAB_ )C D is the CD element of the matrices TAB_ . Consequently, the
complex conjugate matrices T + AB
_ are
C_ D_
_
AB 0_ 0_ 0_ 1_ 1_ 0_ 1_ 1_
D_ _00 "   " :
(TAB
_ ) C_
+
= _ (6.90)
01    
_
10    
1_ 1  
58 CHAPTER 6. GEOMETRY OF ABSOLUTE. . .

Proposition 6.7. In the Carmeli matrices the rst structural Cartan

equations (A) of the A4 geometry have the form
@C D_ Ai B_ @AB_ Ci D_ = (TC D_ )AP Pi B_ + Ai R_ (TDC R_
_ ) B
+
_

P i i
(TAB_ )C P D_ C R_ (TBA R_ _ :
_ ) D
+
(6.91)
Proof. We will write the equations (6.75) as
rk Ci D rk Ai B = TC E k DE
_
i + i T F
_ CF D k _ _
+ _
_

TA C k Ci B_ Ai E_ TB+_Ek_ : (6.92)

It is easily seen that the equations (6.92) represent the dierence of the two
relationships
rk Ci D_ = TC E k DE
i +  i T +F_ ;
_ C F_ D_ k (6.93)
rk Ai B = TAC k Ci B + Ai E TB E k :
_ _ _
+ _
_ (6.94)
Multiplying (6.93) by  k AB_ , and (6.94) by  k C D_ , we get
rk Ci D Ak B = TC E k DE
_ _
i k + i T F k ;
AB C F D k AB
_ _ _
+ _
_ _ (6.95)

rk Ai B Ck D = TAP k Pi B Ck D + Ai E TB E k Ck D :

_ _ _ _ _
+ _
_ _ (6.96)
We now introduce the notation
(TAB_ )C E = TC E k Ak B_ (6.97)
and
@AB_ = Ak B_ rk ; (6.98)
and rewrite the relationships (6.95) and (6.96) as
@C D_ Ai B_ = (TC D_ )A P Pi B_ + Ai R_ (TDC
+ R_ ;
_ ) B _ (6.99)

@AB_ Ci D_ = (TAB_ )C P Pi D_ + Ci R_ (TBA

+ R_ :
_ ) D_ (6.100)
Subtracting from (6.99) the equality (6.100), we will arrive at the rst struc-
tural Cartan equations (6.91) of the A4 geometry, written in terms of Carmeli
matrices.
Consider now the second structural Cartan equations
(B s+ ); written in matrix forms
Rkn + 2r[kTn] [Tk ; Tn] = 0: (6.101)
Multiplying the quantity Rkn by  k AB_ and  n C D_ , we will introduce the
traceless Carmeli matrix
RABC k n
_ D_ = Rkn AB_ C D_ (6.102)
6.4. CARMELI MATRICES. . . 59

with the components [44-46]

   
1 0 ; R 10 00
R010
_ 0_ =
+ 2 _ 0_ = ;
 20 10
100
 2 1  
3 2 2 12 02
R111
_ 0_ =
4  3 ; R110
_ 1_ =
22 ;
12 (6.103)
2 + 11  
R110
_ 0_ =
+ 
1

01
+ ;
 3 21 2 11 
2 + 11 +  
R100
_ 1_ =
3 + 21
1
2 11 
01
:

Proposition 6.8. In terms of Carmeli spinor matrices (6.87) and (6.102),

the second structural Cartan equations (B s+ ) of the A4 geometry become
RABC
_ D_ = @C D
_ TAB
_ @AB_ TC D_ (TC D_ )A F TF B_ (TDC F_ T _ +
_ ) B
+
_ AF

F
+(TAB_ )C TF D_ + (TBA F
_ TC F_ + [TAB _ ; TC D
_ ]: (6.104)
_
_ ) D
+

Proof. We write the equations (6.101) as

Rkn = 2r[nTk] + [Tk ; Tn ] (6.105)
or
Rkn = rnTk rk Tn + Tk Tn TnTk : (6.106)
Multiplying this by  k AB_  n C D_ , we will have
RABC
_ D_ = @C D
k
_ Tk AB_ @AB_ TnCn D_ + TAB_ TC D_ TC D_ TAB_ =
= @C D_ TAB_ @AB TC D_ (@C D_ AB_ k @AB_ C D_ k )Tk +
+[TAB_ ; TC D_ ]: (6.107)
We have used here the condition that
k AB_ Ck E_ = ÆCAÆEB__ (6.108)
and the notation
@AB_ = Ak B_ rk : (6.109)
If now in (6.107) we use the relationships (6.99) and (6.100), we will get the
equations (6.103).
Let us write the second Bianchi identities (D s+) of the A4 geometry in
matrix form  
rn Rkn +[R kn ; T n] = 0: (6.110)
n
Multiplying these equations by  E F_ , we will render them in terms of
Carmeli matrices as follows:
 n (rC D_  AB_ ) 
@ C D_ RE FC
_ D_ +EF_ n RABC_ D _ +
 
+(rk  kC D ) RE FC [T C D ; R E FC _ ] = 0:
_ _
_ D _ _ D (6.111)
60 CHAPTER 6. GEOMETRY OF ABSOLUTE. . .

Using the relationship (6.99), we can rewrite the identities (6.111) as

 
@ C D_ RE FC
_ D_ (T C D_ )AE R AFC _ D _
 
(T +DC )F_ B R E BC _ +(TP
D )CP R _ _ +
_ _ _
_ D E FC D
+C Q _D _  C D_ 
+(TQ_ ) RE FC _ D _ +[T ; R E FC
_ D _ ] = 0: (6.112)

6.5 Component-by-component rendering of

structural equations of A4 geometry
Let us now write the equations (6.91) component by component. For con-
venience, we will introduce the following notation:
AiC DA
_ B_ = @C D
i
_ AB_ @AB_ Ci D_ = (TC D_ )A P Pi B_ +
+Ai R_ (TDC R_ _ (T _ )C P  i
_ ) B
+
AB
i
P D_ C R_ (TBA_ ) D
+ R_ _ : (6.113)
Also, we will denote the components of the spinor derivative as
B_
_ _
@AB_ = A 0 1 ; (6.114)
0 D Æ
1 Æ

and the components of the spinor
-basis as
B_
A 0_ 1_
Ai B_ = : (6.115)
0 li = (Y 0 ; V; Y 2 ; Y 3 ) mi = ( 0; !;  2;  3 )
1 m i = ( ; !;  ;  ) ni = (X o; U; X 2; X 3)
0 2 3

From (6.113), the spinor component Ai000

_ 1_ will be

Ai000 i
_ 1_ = @00_ 01_ @01_ 0i 0_ = (T00_ )0P Pi 1_ + 0i R_ (T00_+ )R_ 1_
(T01_ )0 P Pi 0_ 0i R_ (T10 + R _
_ ) 0 _ (6.116)
or


Ai _ _ = @00_ 0i 1_ @01_ 0i 0_ = (T00_ )00 0i 1_ + (T00_ )0 1 1i 1_ +
 0001 

+ 0i 0_ (T00
_ )
+ 0_
_
1 +  i (T + )1_
01_ _
00 1_ (T01_ )00 0i 0_ + (T01_ )0 1 1i 0_
 
0i 0_ (T10
+ 0_ i
_ ) 0_ + 01_ (T10
_ ) 0_ :
+ 1_
(6.117)
6.5. COMPONENT-BY-COMPONENT RENDERING. . . 61

Using the notation of (6.89)-(6.90) and (6.114)-(6.115) for the components

(TC D_ )A P ; (TBA
+ R_ _ ; @ _ and  i , we will obtain, by (6.117),
_ ) D AB AB_



Dmi Æli = "mi + ( )ni + li + mi ( " )



li + (  )mi li  + mi ( ) =
= ( +   )li ni + m i + ( + " ")mi : (6.118)
Since the vectors mi and li have the following components:
li = (Y 0 ; V; Y 2 ; Y 3 ); mi = ( 0; !;  2;  3 );
it follows from (6.118) that
ÆV D! = ( +   )V + U ! ( + " ")!; (6.119)
ÆY  D  = ( +   )Y  + X    ( + " " )  ; (6.120)
 = 0; 2; 3:
In a similar manner we nd the following component rendering of the rst
structural Cartan equations of the A4 geometry
ÆV D! = ( +   )V + U ! ( + " ")!; (A:1)

ÆY  D  = ( +   )Y  + X   ( + " ")  ; (A:2)

Y  DX  = ( + )Y  + (" + ")X  ( +  )  ( +  )!; (A:3)

V DV = ( + )V + (" + " )U ( +  )! ( +  )!; (A:4)
ÆU
! = V + (   )U + ! + ( + )!; (A:5)

ÆX 
  = Y  + (   )X  +  + ( + )  ; (A:6)
Æ! Æ! = ( )V + ( )U (  )! ( )!; (A:7)
 
Æ  Æ = ( )Y  + ( )X  (  ) ( )  ; (A:8)
 = 0; 2; 3;
and the complex conjugate equations (A:1) (A:8) (all in all 24 independent
equations).
Let us now look at the equations (6.107) and write them componentwise.
For instance, we will derive the R010
_ 0_ component of these equations

R010
_ 0_ = @00_ T01_ @01_ T00_ (T00_ )00 T01_ (T00_ )0 1 T11_ +
+(T0+0_ )0_ 1_ T00_ (T0+0_ )1_ 1_ T01_ + (T01_ )0 0 T00_ + (T01_ )0 1T10_ +
+(T1+0_ )0_ 0_ T00_ + (T1+0_ )1_ 0_ T01_ + T01_ T00_ T00_ T10_ : (6.121)
62 CHAPTER 6. GEOMETRY OF ABSOLUTE. . .

Using the matrices (6.89)-(6.90), (6.103) and the spinor derivative (6.114),
we can represent (6.121) as
     
1 0 =D   Æ " 
2 + 2 
1      " 
"   +   "  +
         "  (6.122)
+"   + "     +
      "  
+ "     +
 "  
     
+   "  "    :
   "  "  
These equations split into the following three independent equations:
(D  + ") (Æ  +  )" ( +  ) + ( + ) 1 = 0;
(D   3" + " ) (Æ  +   3 ) 0 = 0;
(D  + " + " ) (Æ +   +  )  +  2 2 = 0:
Similarly, we will obtain the following independent equations (B s+ ):
(D  " ") (Æ 3  +  )
 +  00 = 0; (B s+ :1)
(D   3" + ") (Æ  +   3 )
0 = 0; (B s+ :2)
(D  " + ") (
3 )   
1 10 = 0; (B s+ :3)
(D  " + 2") (Æ  +  )"  +  + 
 10 = 0; (B s+ :4)
(D + " + ") (
)" ( +  ) ( +  )
 +  +  2 11 = 0; (B s+ :5)
(D  + 3" ") (Æ +  +   )  +  20 = 0; (B s+ :6)
(D  + ") (Æ  +  )" ( +  ) + ( + )
1 = 0; (B s+ :7)
(D  + " + ") (Æ +   +  )  + 
2 2 = 0; (B s+ :8)
(D + 3" + " ) (
+  + )  ( +  )
6.6. CONNECTION OF STRUCTURAL. . . 63

3 21 = 0; (B s+ :9)
(
+  +  + 3 ) (Æ + 3 +  +   ) + 4 = 0; (B s+ :10)
(Æ    ) (Æ 3 +  ) +  ( )+
+ 1 01 = 0; (B s+ :11)
(Æ  + 2 ) (Æ +  )  +  ( )
( )"  + 2 11 = 0; (B s+ :12)
(Æ  + 3 ) (Æ +  +  +  ) ( ) +
+ + 3 21 = 0; (B s+ :13)
(Æ  +  +  ) (
+ + )  +  +
+"  12 = 0; (B s+ :14)
(Æ  + 3 + ) (
+  + + ) +
+ 22 = 0; (B s+ :15)
(Æ   + ) (
+  3 + ) +
+ 02 = 0; (B s+ :16)
(
+  ) (Æ +    ) + 
 + 2 + 2 = 0; (B s+ :17)
(
+ ) (Æ +   ) ( + ") +
+( +  ) + 3 = 0: (B s+ :18)
In addition to these equations, the second structural Cartan equations (B )
include the complex conjugate equations
R+kn + 2r[k Tn+] [Tk+; Tn+] = 0: (B s )
We can write these equations in terms of components by replacing the equations
(B s+:1){(B s+:18) by their complex conjugate equations.
6.6 Connection of structural Cartan equations
of A4 geometry with the NP formalism
In 1962 Newman and Penrose [40] put forward a system of nonlinear spinor
equations, which appeared to be extremely convenient in the search for novel
solutions of Einstein's equations. In the work [47] by the author of this book is
was shown that the equations of the Newman-Penrose formalism coincide with
the structural Cartan equations of the geometry of absolute parallelism. Indeed,
with spinor Carmeli matrices TC D_ one can connect the spintensor TFAC D_ using
the relationships
(TC D_ )A P = TA P k  k C D_ = T P AC D_ = "PF TFAC D_ : (6.123)
64 CHAPTER 6. GEOMETRY OF ABSOLUTE. . .

Using the matrix elements (6.162) of the Carmeli matrices and the funda-
mental spinor  
AB 0 1
" = "AB = 1 0 ;
we will obtain the following notation for the components of the spintensor
TABC D_ :
C D_
AB 00_ 01_ 10_ 11_
TABC D_ = 00     : (6.124)
(01) "  
11    

Proposition 6.9. First structural Cartan equations of the A4 geometry

coincide with the "coordinate equations" [40]
@AB_ Ci D_ @C D_ Ai B_ = "PQ (TPAC D_ QB i TPCAB_ QD i )+

+"R_ S_ (T R_ B_ DC i i
_ AS_ T R_ D_ BA
_ C S_ ) (6.125)
in the Newman-Penrose formalism.
Proof. We will write the structural Cartan equations (A) of the geometry of
absolute parallelism as
@C D_ Ai B_ @AB_ Ci D_ = (TC D_ )AP Pi B_ + Ai R_ (TDC
+ _R
_ ) B _

P i i
(TAB_ )C P D_ C R_ (TBA R_ :
_ ) D (6.126)
+
_

Using the relationship (6.123), we will represent the equations (6.126) as

@C D_ Ai B_ @AB_ Ci D_ = "PQ (TPAC D_ QB i TPCAB_ QD i )+

+ "R_ S_ (T R_ B_ DC i i
_ AS_ T R_ D_ BA
_ C S_ ) :

It is easily seen that these equations are equivalent to (6.125).

We now write the well-known decomposition of the Riemannian tensor Rijkm
into irreducible representations
1
Rijkm = Cijkm 2g[i[k Rm]j ]
Rg g ; (6.127)
3 i[m k]j
where Cijkm is the Weyl tensor (10 independent coordinates); Rij is the Ricci
tensor (nine independent coordinates); R is the scalar curvature. The spinor
representation of these quantities using the Newman-Penrose formalism looks
like [48]
Cijkm $ ABCD "A_ B_ "C_ D_ + "AB "CD A_ B_ C_ D_ ; (6.128)
Rij $ 2AB A_ B_ + 6"AB "A_ B_ ; (6.129)
R = 24; (6.130)
6.6. CONNECTION OF STRUCTURAL. . . 65

where spinors ABCD and AB A_ B_ have the following symmetry properties:
ABCD = (ABCD) ; AB A_ B_ = (AB )A_ B_ : (6.131)

By denition the spinors ABCD and AB A_ B_ are transformed following the
D (2:0) and D (1:1) irreducible representation of the groups SL+ (2:C ), respec-
tively.
If we now put in juxtaposition to the Riemann tensor Rijkm a spintensor
following the rule
Rijkm $ RAAB
_ BC
_ CD _;
_ D

then in terms of the spinors (6.128)-(6.130) it can be written as

RAAB
_ BC
_ CD _ = ABCD "A_ B_ "C_ D
_ D _ + "AB "CD A_ B _ +
_ C_ D

+AB C_ D_ "CD "A_ B_ + CDA_ B_ "AB "C_ D_ +

+2("AC "BD "A_ D_ "C_ D_ +
+"AB "CD "A_ D_ "B_ C_ ): (6.132)

This spintensor being skew-symmetric in the pair of indices AA_ and B B_ , we

will write it as
1
RAEC
_ BD
_ PF
_ Q_ = ("E_ B_ RACDP_ F Q_ + "AC RE_ BD _ );
_ P_ F Q (6.133)
2
where
1
RACDP_ F Q_ = "E_ B_ RAEC
_ BD _;
_ P_ F Q (6.134)
2
1 AC
RE_ BD _ =
_ P_ F Q " RAEC_ BD _:
_ P_ F Q (6.135)
2
Substituting into these relationships the equality (6.132) gives
RACDP_ F Q_ = ACDF "P_ Q_ + AC Q_ P_ "FD + "P_ Q_ ("CD "AF + "AD "CF ); (6.136)

RE_ BD
_ BPQ
_ = "DP E_ B_ P_ Q_ +  B_ EPD
_ "Q_ P_ + "DP ("B_ P_ "E_ Q_ + "E_ P_ "B_ Q_ ): (6.137)

Proposition 6.10. The second structural Cartan equations (B s+) are

equivalent to the equations [40]
ACDF "E_ B_ + AC B_ E_ "FD + "E_ B_ ("CD "AF +
+"AD "CF ) @DB_ TACF E_ + @F E_ TACDB_ +
PQ
+" (TAPDB_ TQCF E_ + TACP B_ TQDF E_ TAPF E_ TQCDB_
TACP E_ TQFDB_ ) +
+"R_ S_ (TACDR_ T S_ B_ EF
_ TACF R_ T S_ E_ BD
_ ) = 0 (6.138)
66 CHAPTER 6. GEOMETRY OF ABSOLUTE. . .

in the Newman-Penrose formalism.

Proof. We write the equations (B s+ ) in terms of the Carmeli matrices
RF ED
_ B_ = @DB
_ TF E
_ @F E_ TDB_ (TDB_ )F S TS B_
(TED F_ T _ + (T _ ) S T _ +
_ ) B
+
_ FF F E D SB
+(TEF F
_ TDF_ + [TF E
_ ; TDB_ ]:
_
_ ) B (6.139)
+

Using the relationships (6.123), we can represent the equations (6.139) as

RACF ED
_ B_ @DB_ TACE F_ + @E F_ TACDB_ + T S FDB_ TACS E_ +
+T F E_ BD T S DF E_ TACS B_ T F B_ EF
_ _
_ TACF F_ _ TACDF_ +
PQ
+" (TAPDB_ TQCF E_ TAPF E_ TQCDB_ ) = 0;
or as
RACF ED
_ B_ @DB_ TACE F_ + @E F_ TACDB_ + "PQ (TAPDB_ TQCF E_ +
+TACP B_ TQDF E_ TAPF E_ TQCDB_ TACP E_ TQFDB_ )+ (6.140)
+"R_ S_ (TACDR_ T S_ B_ EF
_ TACF R_ T S_ E_ BD
_ ) = 0;

where we have introduced the spinor indices in the matrices RABC

_ D_ and TAB
_
following the rule
_ ! REFABC
RABC k n ;
_ D _ D_ = REFkn AB_ C D
_
(6.141)
TAB_ ! TCDAB_ = TCDk AB_ :
k

Substituting into (6.140) the relationship (6.136), we will arrive at the equa-
tions (6.138).
Spintensors ABCE and AB C_ E_ have the following notation for their com-
ponents [38]:
CE
AB 00 01 11
ABCE = 00 0 1 2 ; (6.142)
01 3
11 4
C_ E_
AB 0_ 0_ 0_ 1_ 1_ 1_
AB C_ E_ = 00 00 01 02 ; (6.143)
01 10 11 12
11 20 21 22
 = : (6.144)
Using the relationships (6.114), (6.115), (6.124), we can expand the equations
(6.126) of the Newman-Penrose formalism component by component to arrive
at the equations (A:1) (A:8) plus the complex conjugate equations. Using
6.6. CONNECTION OF STRUCTURAL. . . 67

the relationships (6.142)-(6.144) and (6.114) we also can expand the equations
(6.138) of the Newman-Penrose formalism componentwise. We will thus end up
with the equations (B s+:1){(B s+:18).
The spinor counterpart of the dual Riemann tensor
 1
Rijkm = 2 "km sp Rijsp (6.145)

can be written as

RAAB
_ BC
_ CD
_ D_ = i "AB "CD A_ B _ C_ D
_ ABCD "A_ B_ "C_ D_
 CDA_ B_ "AB "C_ D_ + AB C_ D_ "CD "A_ B_ +
+2("AC "BD "A_ B_ "D_ C_ + "AB "CD "A_ C_ "B_ D_ )) : (6.146)
It follows that
 1
RABC
_ DEF
_ = "P_ Q_ RABC_ DE
_ PF
_ Q _ = i ( "B_ D_ ACEF +
2 _

+"EF AC B_ D_ + "B_ D_ ("AE "CF + "CE "AF )) ; (6.147)

also
 
1
RABC
_ D _=
_ P_ Q "EF RABC
_ DE
_ PF
_ Q_ = i "A_ C_ B
_D_ P_ Q
2 _

"P_ Q_  AC B_ D_ + "AC ("D_ P_ "B_ Q_ + "B_ P_ "D_ Q_ ) : (6.148)

The dual Weyl tensor C ijkm corresponds to the spintensor of the form
 
C ijkm $C AAB
_ BC
_ CD _ = i("AB "CD A_ B
_ D _ C_ D
_ ABCD "A_ B_ "C_ D_ ):
The self-dual spintensor RABC
_ DEF
_ will be

RABC
_ DEF
_ = i RABC
_ DEF
_ = ACEF "B_ D_ ; (6.149)
whereas the anti-self-dual tensor is

R ABC
_ D _ = i RABC
_ P_ Q _ D _ = "AC E
_ P_ Q _:
_ B_ P_ Q (6.150)

Proposition 6.11. The second Bianchi identities (D s+) of the A4 geo-

metry in the spinor
-basis can be represented as
1
" F EG_ HR
_ X_
@P X_ RABGHF ABCR T R F F D_
2i C D_ _ E _

3 RPB (A TC ) RP D_ + RB D_ X_ TA R C X_ +
X_ X_
+AB X_ E_ T D_ E_ C + AB D_ X_ T E_ E_ C = 0; (6.151)
68 CHAPTER 6. GEOMETRY OF ABSOLUTE. . .

where
"C DF
_ EG
_ HR
_ X_
= i("CG "RF "D_ E_ "H_ X_ "CF "GR "D_ H_ "E_ X_ ): (6.152)
Proof. We will write the equations (6.79) as
  
rn RACkn REAkn TC En RAEkn TC En = 0: (6.153)

Multiplying these equations by Ck D_ gives

 
@ F E_ R BAC DF _ E_ r n
n F E_ +
_ E_ + R BAC DF
 k F E_  RS_
+ RBARSF_ E_ C D_ @ k
 PF E_  PF E_ = 0:
_ E_ TA
RBPC DF RPAC DF
_ E_ TB (6.154)
Here we have used the relationships (6.94) and (6.133). Substituting into
(6.154) the relationship (6.148), we will get
 
iDABC D_ + AF EP i  P RF
_ E
2i P R RBAC D_ F E = 0;
iC D_ RBA
_ _ _
_ R_

where AF EP i
_ R_ stands for the equations (6.125), rewritten as

AABC i i _ @C D_  iAB_
_ D_ = @AB
_  CD
 
"PQ TPAC D_  iQB_ TPCAB_  iQD_
i i

"R_ S_ T R_ B_ DC
_  AS_ T R_ D_ BA
_  C S_ = 0; (6.155)
and DABC D_ = 0 denes the equations (6.151)
1
DABC D_ = "C D_ F EG
_ HR
_ X_
@RX_ RABGHF _ E
2i _

ABCR T RF F D_ 3 RPB (A TC ) RP D_ + RB D_ X_ TA R C X_ +

+AB X_ E_ T X D_ E_ C + AB D_ X_ T X E_ E_ C = 0;
_ _

which proves the Proposition.

Proposition 6.12. The second Bianchi identities (6.151) of the A4 geo-
metry coincide with the Bianchi identities in the work by Newman-Penrose [40].

@ P D_ ABPC @ X (C AB )D_ X_ 3 PR(AB TC ) PR D_

ABCP T P R R D_ + 2T P (AB X_ C )P X_ D_
T X_ D_ V_ (A BC ) X_ V_ T X_ V_ V_ (A BC ) X_ D_ = 0; (6.156)
6.6. CONNECTION OF STRUCTURAL. . . 69

3@AB_  + @ P X AP B_ X_ "V_ W_ AP X_ W_ T B_ X_ V_ P +
_


+AP B_ X_ T X_ W_ V_ P + PRB_ X_ TA PR X_ +
+AP B_ X_ T P R R X_ = 0 (6.157)
Proof. Using (6.147) and the equality
 1
RABC D_ RX = 2 "C D_ F EGHRX RABGH_ F E_ ;
_ _ _ _

we nd that in (6.151)
1
"C D_ F EG
_ HR
_ X RX_ =
@P X_ RABGHF
_ E_ = @P X
_ RABC D
_

 2 i _

= @P X_ "D_ X_ ABC R "AB C R D_ X_ "D_ X_ ("CA "B R + "BA "C R ) :

Substituting this relationship into (6.151) gives

@ P D_ ABPC @C X_ AB D_ X_ + 2"C (A @B )D_  ABCR T RF F D_
X_
3 RPB (A TC ) RP D_ + RB D_ X_ TA R C X_ + AB X_ E_ T D_ E C +
_

+AB D_ X_ T X E_ E_ C = 0:
_
(6.158)
The part of (6.158) symmetrical in the indices C and B can be written as
(6.156); and the part skew-symmetrical in these indices looks like (6.157).
By writing the second Bianchi identities (D s+) of the A4 geometry compo-
nent by component, we obtain [40]
(D 4 2") 1 (Æ 4 +  ) 0 +
+3 2 + (Æ 2 2 +  )00
(D 2 2")01 211 +
+2 10  02 = 0; (D s+ :1)

(D
3) 2 (Æ + 2 2) 1 +
+2 3 +  0 + (Æ 2 +
+ )10 (D 2)11 21
 12 00 +  01 +  20 D  = 0; (D s+ :2)

(D 2 + 2") 3 (Æ + 3 ) 2 +
+2 1 +  4 + (Æ 2 + 2 +
+ )20 (D 2 + 2")21
210 + 2 11  22 2Æ  = 0; (D s+ :3)
70 CHAPTER 6. GEOMETRY OF ABSOLUTE. . .

(Æ 4 2 ) 1 (
4 + ) 0 +
+3 2 + (Æ 2 + 2 )01 (D 2" + 2"
 )02 212 + 2 11
00 = 0; (D s+:4)

(Æ 3 ) 2 (
+ 2 2 ) 1 + 2 3 +
+ 0 + (Æ + 2 )11 (D + 2"
 )12 22 01 +  02 +
+ 21 10 Æ  = 0; (D s+:5)

(Æ + 2 2 ) 3 (
+ 3) 2 + 2 1 +
+ 4 + (Æ + 2 + 2 )21
(D + 2" + 2" )22
211 + 2 12 20 2
 = 0; (D s+:6)

(D + 4" ) 4 (Æ + 4 + 2) 3 +
+3 2 + (
+ 2 2 + )20
(Æ + 2 2 )21 2 10 +
+211  22 = 0; (D s+:7)

(Æ + 4  ) 4 (
+ 2 + 4) 3 + 3 2 +
+(
+ 2 + 2)21 (Æ + 2 +
+2  )22 2 11 + 212
 20 = 0; (D s+:8)

(D 2 2)11 (Æ 2 2 +
+ )10 (Æ 2 2 +  )01 +
+(
+ 2 2 +  + )00 +
+ 12 + 21  02
 20 + 3D  = 0; (D s+:9)

(D 2 + 2" )12
(Æ + 2 2 )11 (Æ + 2
2  +  )02 + (
+ 2 2 +
+)01 + 22  00
10  21 + 3Æ  = 0; (D s+:10)
6.7. VARIATIONAL PRINCIPLE. . . 71

(D + 2" + 2"  )22

(Æ + 2 + 2  )21 (Æ +
+2 + 2  )12 + (
+ 2 +
+2 )11  10  01 +
+ 20 + 02 + 3
 = 0: (D s+ :11)
To arrive at the complete set of the second Bianchi (D ) identities of the
A4 geometry, we will have to add to these equations their complex conjugate
(D s ).

6.7 Variational principle of derivation of the

structural Cartan equations and the second
Bianchi identities of A4 geometry
To begin with, we will consider the derivation of the structural equations (B )
and of the second Bianchi identities (D ) for self-dual and anti-self-dual elds of
Riemannian curvature, whose Carmeli matrices obey the conditions

Rkn = i Rkn ;

R+kn = i Rkn ;
where
Rkn + 2r[k Tn] [Tk ; Tn] = 0;
R+kn + 2r[k Tn+] [Tk+; Tn+] = 0;
and
 1
Rkn = 2 "knps Rps ;
 1
R +kn = 2 "knps R+ps :
Let us take the Lagrange function in the form
L1 = 1 ( g )1=2T r (RknRkn ) + complex conjugate part: (6.159)
4
Varying this expression in Tk and Tk+ , we will arrive at the equations (D )
rn R kn +[R kn ; T n] = 0; (6.160)
 
rn R +kn + [R +kn ; T +n ] = 0: (6.161)
For arbitrary elds of Riemannian curvature the Lagrange function looks
like
 
1  kn 1
L2 = 2 ( g ) T r R ( 2 Rkn 2r[k Tn] + [Tk ; Tn]) + c.c. part: (6.162)
1=2
72 CHAPTER 6. GEOMETRY OF ABSOLUTE. . .

 
Variation of this Lagrangian in R kn and R+ kn yields the second Bianchi iden-
tities (D)
 
rn R kn +[Rkn ; T n] = 0; (D s+ )
 
rn R+ kn +[R +kn ; T +n ] = 0: (D s )
On the other hand, variation of the Lagrangian (6.162) in Tk and T + k gives
the second structural Cartan equations (B ) of the A4 geometry

Rkn + 2r[kTn] [Tk ; Tn] = 0; (B s+ )

R+kn + 2r[kTn+] [Tk+ ; Tn+] = 0; (B s )
and
 1
Rkn = 2 "knps Rps ;
 1
R +kn = 2 "knps R+ps:
Independent variables in the Lagrangian (6.162) are the quantities Rkn , R+kn ,
Tk , and Tk+ . To obtain from them using the variational principle, the rst
structural Cartan equations (A) of the A4 geometry
r k i
[
]
T[k  i]  [iTk+] = 0; (As)
we will have to introduce into the Lagrangian (6.162) as independent variables
the matrices  i . This can be done by modifying the Lagrangian (6.162) as it
has been done in [49].
We now write the equations (A); (B ) and (D ) in spinor form :
(A) Ai ABC
_ D_ = 0; (6.163)
(B ) BF EACD
_ B_ = 0 + c.c. equations; (6.164)
(D ) DABC D_ = 0 + c.c. equations; (6.165)
where
AiABC
_ D_ = @AB
i
_ C D
_ @C D_ Ai B_ "PQ (TPAC D_ QB i
i ) "R_ S_ (T
TPCAB_ QD i
R_ B_ D_ C AS_
T R_ D_ BA i
_ C S_ ) = 0; (6.166)

BACF ED
_ B_ = RACF ED
_ B_ @DB_ TACE F_ + @E F_ TACDB_ +
+"PQ (TAPDB_ TQCF E_ +
+TACP B_ TQDF E_ TAPF E_ TQCDB_ TACP E_ TQFDB_ ) +
+"R_ S_ (TACDR_ T S_ B_ EF
_ TACF R_ T S_ E_ BD
_ ) = 0; (6.167)
6.7. VARIATIONAL PRINCIPLE. . . 73

DABC D_ = 1 "C D_ F EG
_ HR
_ X
_
@RX_ RABGHF
_ E_
2i
R F
ABCR T F D_ 3 RPB (A TC ) D_ + RP

+RB D_ X_ TAR C X_ + AB X_ E_ T X D_ E_ C + AB D_ X_ T X E_ E_ C = 0

_ _
(6.168)
and consider the Lagrangian
 
 1
L3 =R B Q_ AQkn (2rnTABk + 2TPAn TB P k ) R P + c.c. part:
_

4 BPA nk
(6.169)
 B AQkn nkjm B A Q nkjm
=" R Q_ jm and "
_ _
Here R Q_ is a completely skew-symmetrical
Levi-Chivita symbol.
If we take RB Q_ AQkn
_
and TPAn to be independent variables and use the
conventional variational procedure, we will obtain the following equations:
1
(B s+ ) R P_ 2r[k TjAB jn] + 2TPA[k T P jB jn] = 0; (6.170)
2 B P_ A kn
(B s ) complex conjugate equations; (6.171)
rk R B QA 
(D s+) _
Q_ nk 2 R Q_ Pk
P Q_ (A jnkj TB ) = 0; (6.172)
(D s ) complex conjugate equations: (6.173)
Multiplying equations (6.170) by C D_ n F E_ n gives
@F E_ TABC D_ P
@C D_ TABF E_ + TPAF E_ TBA P
D_ TP AC D_ TBF E_
1
R _ P F EC _ + TABn (@C D
_ F E_
n @F E_ C D_ n ) = 0: (6.174)
2 B QA _ D

Using the notation (6.166) and (6.167), we will write (6.174) as

BACF ED n
_ B_ + A CDF E_ TABn = 0: (6.175)
We will now multiply (6.172) by C D_ k to get the relationship
 
Q_ _ _ + R _ Q_ _ _ rk  F E_ +
@ F E_ R B QA
_ C DE F B QA C DE F k
 Q_ _ _  n _ @ F E_  RS_
+ RB QA
_ RSE F C D n
 _ _ 
Q _ _ T PF E R _ Q _ _ T PF E = 0
_ _
RB QP
_ C DE F A P QA C DE F B (6.176)
or, from (6.166) and (6.167),
 
n 1  QP
_ RE
_ F_ P R_  Q_ E F_
iDABC D_ + AF EP
_ R_ 
2 nC D_ RB QA
_ n RB QA
_ C D_ = 0:
(6.177)
74 CHAPTER 6. GEOMETRY OF ABSOLUTE. . .

Here we have also used the relationship

 Q_ _ _ = i (2
RB QA
_ C DE F BACF "DE_ 2"CF AB D_ E_ + 2"D_ E_ ("BF "AC +
+"BC "AF )) :
It is clear that from the Lagrangian (6.169) it is impossible to obtain the rst
structural Cartan equations (A) of the A4 geometry, since it does not contain
C D_ n.
Let us add to the Lagrangian (6.169) the term
Aj BC
_ D_
AjABC
_ D_ (6.178)

where the quantities ABC

_ D_
play the role of Lagrange factors
L4 = L3 + Aj BC
_ D_
AjABC
_ D_ + c.c. part: (6.179)

The quantities Aj BC , just like AjABC

_ D_
_ D_ , are Hermitian matrices, which are

skew-symmetrical in the pair of indices [49] AB_ and C D_ . Varying the Lagrange
density (6.179) in C D_ n gives [49]
AjABC
_ D_ = 0 (6.180)
and
DABC D_ = kP R_ nB
X_ (n
AXP
_ R
_ nAXP
_ R_ )
k
C D_ = 0: (6.181)
Since nAXP
_ R_ are Hermitian matrices, from (6.181) we have the equations
(D s+ )
DABC D_ = 0: (6.182)
Hence varying the complex conjugate part of the Lagrangian (6.179) gives
D A_ B_ CD
_ = 0: (6.183)
and of the Lagrangian (6.179) in RB QA
_
Qkn
_
gives
BACF ED n
_ B_ + A CDF E_ TABn = 0 (6.184)
or, from (6.180),
BACF ED
_ B_ = 0: (6.185)
Variation of the complex conjugate part in R B Q AQkn
_
_
yields
B ACF ED
_ B_ = 0: (6.186)
It has thus been shown that from the Lagrangian (6.179) follow the rst and
second of the structural Cartan equations of the A4 geometry (equations (6.180),
(6.185) and (6.186)), and also the second of the Bianchi identities (equations
(6.182) and (6.183)).
6.8. DECOMPOSITION OF SPINOR. . . 75

6.8 Decomposition of spinor elds of A4

geometry into irreducible parts
The torsion tensor
::i
jk of the A4 space has 24 independent components, and
it can be represented as the sum of three irreducible parts as follows:
2 1

i:jk = Æ[ik
j ] + "njks
^ s +
i:jk ; (6.187)
3 3
where

i:jk = g imgks
::s
mj ; (6.188)
i
and the vector
j , pseudovector
^ j and the traceless part of the torsion
:jk
are given by

j =
i:ji; (6.189)
1

^ j = "jins
ins; (6.190)
2
s

:js = 0;
ijs +
jsi +
sij = 0: (6.191)
In the spinor basis the spinor representation of the Ricci rotation coeÆcients
TABC C_ has the form [40]
 
1 1
TABC C_ = AABC C_ + ("AC B C_ + "BC AC_ ) ; (6.192)
2 3
where the spinor AABC C_ is completely symmetrical in the unprimed indices
AABC C_ = A(ABC )C_ ; (6.193)
and the spinor B C_ is given by

AC_ = AAB B C_ : (6.194)

In turn, the spinor AC_ can be decomposed into the Hermitian and anti-
Hermitian parts:
AC_ = AC_ iAC_ ; (6.195)
where
1 1
AC_ = (AC_ + A_ C ); AC_ = i(AC_ A_ C ) (6.196)
2 2
and
AC_ =  AC_ = C A_ ; AC_ = AC_ = C A_ : (6.197)
The irreducible parts of torsion (6.189)-(6.191) and the spinors (6.193)-
(6.197) are related by

j ! AC_ ; (6.198)

^ j ! AC_ ; (6.199)
76 CHAPTER 6. GEOMETRY OF ABSOLUTE. . .

k

:js ! AABC C : _ (6.200)
Since

ijk = gsk
::s
ij ; (6.201)
we have

AAB_ BC
_ C_ !
ijk ; (6.202)
1

AAB _ C_ =
_ BC (
" +
A_ B_ CC _ "AB ); (6.203)
2 ABC C_ A_ B_

ABC C_ = AC (AB )C_ + C_ (A "B )C : (6.204)
By denition, the spinor AABC C_ is transformed in the D (3=2:1=2) irre-
ducible representation of the group SL(2; C ). Consequently, the spinors AC_
and AC_ are transformed in the D (1=2:1=2) irreducible representation of the
group SL(2:C ). Using the relationship (6.124), we can nd the components of
the spinors AC_ and AC_ [50]
 
1
( + ) 12 (" + " ) 1
( +  ) + 12 (  )
AC_ = 1
2
(  ) + 12 (  )
2
1
( + ) 12 ( + )
; (6.205)
2 2

 
(  ) 12 (" ") 21 (  ) 12 (  )
1
AC_ = i (  ) + 12 (  ) 12 ( ) 12 ( ) :
2
1 (6.206)
2

The Riemann tensor represented in terms of irreducible parts is

1
Rijkm = Cijkm + gi[kRm]j + gj [k Rm]i + 3 Rgi[mgk]j : (6.207)

In the spinor basis this becomes [40]

RAAB _ BC
_ CD _ D_ = ABCD "A_ B_ "C_ D _ +

+"AB "CD A_ B_ C_ D_ + AB C_ D_ "CD "A_ B_ +

+CDA_ B_ "AB "C_ D_ + 2("AC "BD "A_ B_ "C_ D_ +
+"AB "CD "A_ D_ "B_ C_ ): (6.208)
We also have the following connection:
Cijkm ! ABCD "A_ B_ "C_ D_ + "AB "CD A_ B_ C_ D_ ;
Rij ! 2AB C_ D_ + 6"AB "C_ D_ ; (6.209)
R ! 24;
where the spinors ABCD , AB C_ D_ and  meet the following symmetry condi-
tions:
ABCD = (ABCD) ; AB C_ D_ = (AB )(C_ D_ ) ;  = 
and belong to the D (2:0), D (1:1) and D (0:0) irreducible representations of the
group SL(2:C ), respectively.
6.9. SPINOR REPRESENTATION. . . 77

6.9 Spinor set of Einstein-Yang-Mills equations

In the rst part of the book it was shown that the structural Cartan equations
of the geometry of absolute parallelism (A) and (B ) can be represented as an
extended set of Einstein-Yang-Mills equations
r k ea j
[ + T[ikj ]ea i = 0;
] (A)
Rjm gjm R = Tjm ;1
2 (B:1) (6.210)
C i jkm + 2r[k T ijj jm] + 2Tsi[k T sjj jm] = J i jkm : (B:2)
We will write this set of equations in the spinor basis. To this end, we will
make use of the Carmeli matrices and the Newman-Penrose spinor formalism.
Suppose now we have the right spin A4 geometry, then its equations (A) and
(B ) have the form
@C D_ Ai B_ @AB_ Ci D_ = (TC D_ )PA Pi B_ +
+Ai R_ (TDC R_
_ )B
+
_ (TAB_ )PC Pi D_ Ci R_ (TBA R_ ;
_ )D
+
_ (As+)

RABC _ D_ = @C D_ TAB_ @AB_ TC D_ (TC D_ )FA TF B_

(TDC F_ T _ + (TAB_ )FC TF D_ + (TBA F_ T _ + [T _ ; T _ ]; (B s+ )
_ )B _ )D
+ +
_ AF _ CF AB C D
where the components of the matrices Ai B_ , TAB_ and RABC
_ D_ are given by
(6.115), (6.88) and (6.103), respectively.
Proposition 6.13. Equations (B:1) in the spinor basis are
2AB C_ D_ + "AB "C_ D_ = TACB
_ D_: (6.211)
Proof. In terms of the irreducible spinors (6.209) P Q the components of the
spinor matrices RABC
_ D_ are given by [51]
 
(RABC
_ D_ )P
Q="
D_ B_ CAP
Q ("PC ÆAQ + "PA ÆCQ) + "CA P Q D_ B_ ; (6.212)
where
(CABC Q=" Q
_ )P
_ D D_ B_ CAP (6.213)
are the P Q components of the spinor matrices of the Weyl tensor with the
the components
   
1 0 3 2
C010
_ 0_ =
2 1 ; C111
_ 0_ =
4 3 ;
   
2 1 2 1
C110
_ 0_ =
3 2 ; C100
_ 1_ =
3 2 ; (6.214)

and related with the spinor "D_ B_ ("PC Æ AQ + "PA ÆC Q ) and "CA P QD_ B_ are the
trace and traceless parts of the Ricci tensor
1 k
"AB "C_ D_ =   n Rg ; (6.215)
4 AC_ B D_ kn
78 CHAPTER 6. GEOMETRY OF ABSOLUTE. . .

 
1 1
AB C_ D_ =  k AC_  n B D_ Rkn g R : (6.216)
2 4 kn
Substituting relationships (6.215) and (6.216) into (6.211) and multiplying
the resultant expression by  AC_ k  B D_ n , we arrive at the equations (B:1):
We now represent the matrix RABC _ D _ as the sum

RABC
_ D_ = CABC
_ D_ + JABC
_ D_; (6.217)
where the matrix current JABC
_ D_ has the components [52]:

   
J010 1 0
0 ; J111 1 0 1
T ;
_ 0_ = _ =
6
2 1
6
T 0 _ 0
2 0 0
 
J100 1 T100_ 0_ T000 ;
_ 0_ =
_ 0_

2 T101_ 0_ T100
_ 0_
 
1 T011_ 1_ T010
J110
_ 1_ =
2 T111_ 1_
_ 1_
T011 ; (6.218)
_ 1_
   
1 T110 T010 1 1
T 0
J110
_ 0_ =
_ 0_
T101
_ 0_
T110
6
0 T ;
2 _ 1_ _ 0_ 2 1
6
   
J100 1 T110 T010 1 1
T 0 :
_ 1_ = +
_ 0_ _ 0_ 6
2 T101
_ 1_ T110
_ 0_ 2 0 1
6
T
Here
TABC k n T ;
_ =  AC
_ D _ B D_ kn (6.219)
T = g jmTjm; (6.220)
and the energy-momentum tensor Tkn is given in terms of the Ricci rotation
coeÆcients by
2n
Tjm =  r[iTjijjm] + Tsi[iTjsj jm]
 
1 pn
g gjm r[ijTjpjn] + Ts[iTjpjn] :
i i s (6.221)
2
In the special case where the eld T ijk is skew-symmetric in all the three
indices, the tensor (6.219) is [33]
 
Tjm = 1
^ j
^ m 1
g
^ i
^ : (6.222)
 2 jm i
Multiplying this by  j AC_  m B D_ and using (6.199), we get
 
1 1 P Q_ :
TABC
_ D_ =
 AB_ C D_ 2
" AC " B _ P Q
_D _ (6.223)
6.10. FORMALISM OF TWO-COMPONENT. . . 79

In addition, we obtain

T = g jmTjm = 1
^ j
^ j = 1 P Q_ P Q_ : (6.224)
 
Hence the \density of spinor matter" is
1
= P Q_ P Q_ : (6.225)
c2
We substitute (6:217) into the spinor equations (B s+) go get
2AB C_ D_ + "AB "C_ D_ = TACB
_ D_; (B s+ :1)
CABC
_ D_ @C D_ TAB_ + @AB_ TC D_ + (TC D_ )FA TF B_ + (TDC F
_ )B_ TAF_
+

(TAB_ )FC TF D_ (TBA

+ F_ T _ [T _ ; T _ ] = J _ _ :
_ )D_ CF AB C D ABC D (B s+ :2)
To conclude, we will write the extended set of Einstein-Yang-Mills equations
as

@C D_ Ai B_@AB_ Ci D_ = (TC D_ )PA Pi B_ + Ai R_ (TDC

_ )B
+ R_
_

(TAB_ )PC Pi D_ Ci R_ (TBA + R;

_ )D
_
_ (As )
2AB C_ D_ + "AB "C_ D_ = TACB _ D _;
s
(B + :1)
CABC @ T @ T T F F_ T _
+ + ( C D_ A F B_ (TDC
) T + _ )B
+
_ D_ C D_ AB_ AB_ C D_ _ AF

(TAB_ )FC TF D_ (TBA +

_ ) F_ T _ [T _ ; T _ ] = J _ _ :
D_ C F AB C D ABC D (B s+ :2)

_ B;
where the spinor indices take on the values A; B; D : : : = 0; 1, A; _ D_ : : : = 0_ ; 1_ .
6.10 Formalism of two-component spinors
We will introduce the two-component spinors o and i [53], connected with
the components of the spinor dyad  as follows:

0 = o ; 1 =  ; 0__ = o_ ;
1__ = _ ; (6.226)
_ _ : : : = 0_ ; 1_ :
;  : : : = 0; 1; ;
From the orthogonality condition for the spinor dyad
0 1 = 1;
0 0 = 0 0 = 0; (6.227)
1 1 = 0:

0 0 1 0 = Æ ;
0 1 1 0 = " ; (6.228)
80 CHAPTER 6. GEOMETRY OF ABSOLUTE. . .

where  
0 1
" = " = " _ Æ_ = " _ Æ_ =
1 0 ; (6.229)
we derive the normalization condition for the two-component spinors
o  =  o = 1;
o o  
 = o o = 0;   = 0; (6.230)
and also the relationships
" = o   o ; " = o  o  ; " = o   o :
Spinors o and i dene the components of the Newman-Penrose symbols
(6.6)
Ai B_ = i _ A  B_
_
(6.231)
as follows:
0i 0_ = i _ o o_ = li ; 1i 1_ = i _  _ = ni ;
(6.232)
0i 1_ = i _ o _ = mi ; 1i 0_ = i _  o_ = mi :
The vectors li , ni , mi and m i form an isotropic tetrad. The conventional
tetrad ei a can be made up of the vectors of an isotropic tetrad using the rela-
tionships
ei0 = (2) =
(li + ni ) = (2) 1=2 i _ (o o_ +  _ );
1 2

ei1 = (2) 1=2 (mi + m i ) = (2) 1=2 i _ (o _ +  o_ );

ei2 = (2) =
1 2
i(mi m i ) = (2) =
1 2
i _ (o _  o_ );
ei3 = (2) =
1 2
(li ni ) = (2) =
1 2
i _ (oo_  _ ): (6.233)
Using the relationships

TACk = 1 "B_ D_ Ci D_ rk ABi_ ; (6.234)

2
r = i  ri
_ _ (6.235)
we nd the following expressions for the components of the Carmeli matrices
[54]:
 = o o_ o r_ o ;  =  o_  r_  ;
 =  o_ o r_ o ;  = o o_  r_  ;
 = o _ o r_ o ; " = o o_  r_ o ;
(6.236)
 =  _ o r_ o ;  = o _  r_ o ;
 =  _  r_  ; =  _ o r_  ;
 = o _  r_  ;  =  o_ o r_  ;
6.10. FORMALISM OF TWO-COMPONENT. . . 81

0 = Æ o o o oÆ ; 1 = Æ o o o Æ ; (6.237)
2 = Æ o o  Æ ; 3 = Æ o   Æ ;
4 = Æ    Æ ;

00 = 00 =  _ Æ_ o o o_ oÆ_ ; 01 = 10 =  _ Æ_ o o o_ Æ_ ;

02 =  2 =  _ Æ_ o o _ Æ_ ; 11 = 11 =  _ Æ_ o  o_ Æ_ ; (6.238)
12 = 21 =  _ Æ_ o  _ Æ_ ; 22 = 22 =  _ Æ_   _ Æ_ :
It follows from (6.236) that

r o = o o o o o  o  o + "o  
_ _ _ _ _
(6.239)
 o o_ +  o _ +   o_   _ ;

r  = o o o oo  o  o + o  
_ _ _ _ _
(6.240)
 o o_ +  o _ +   o_ "  _ :
The components of the spinor derivative (6.114) can be represented in terms
of two-component spinors as
D = o o_ r_ ;
=  _ r_ ;
(6.241)
Æ = o _ r_ ; Æ =  o_ r_ :
In the formalism of two-component spinors there exists the so-called mod-
ied formalism [53] that takes into account the "primed" symmetry of spinor
quantities. This symmetry allows the replacement
o ! i ;  ! io; (6.242)
o_ ! i _ ; _ ! io_ ;
where the unprimed quantities are replaced by primed ones following the rule
(li )0 = ni ; (mi )0 = m i; (mi )0 = mi ; (ni)0 = li ; (6.243)
 = 0 ;  = o0 ;  = 0 ; (6.244)
 =  0 ;  =  0 ; = "0 :
This symmetry property makes it possible to replace in (6.236) unprimed
quantities by primes ones
 = o_ o o r_ o ;  0 = _ o  r_  ;
 =  o_ o r _ o ;  0 = o o_  r _  ;
 
 = o _ o r_ o ; " = o o_  r_ o ;
(6.245)
 =  _ o r_ o ;  = o _  r_ o ;
0 =  _  r_  ; "0 =  _ o r_  ;
0 = o _  r_  ;  0 =  o_ o r_  ;
82 CHAPTER 6. GEOMETRY OF ABSOLUTE. . .

Here instead of 12 spinor coeÆcients we have only six.

The most general transformation under which spinors o , i and the condi-
tions (6.230) are retained is
o ! Co ;  ! C 1  ; (6.246)
where C is a complex transformation that forms a subgroup of boosts and
three-dimensional rotations. The components of the isotropic tetrad (6.232)
vary under these transformations as follows:
li ! A 1 li ; ni ! Ani ; mi ! ei mi ;
A = CC; ei = CC 1 : (6.247)
We will now dene a scalar quantity with the following properties:

 ! C P C q : (6.248)
This quantity is said to be a spin and boost weight scalar of the type (p; q )
[53]. It follows from (6.246) that the components of the spinors o and i are
scalars of types (1; 0) and ( 1; 0), respectively. The components of the isotropic
tetrad will be
li : (1; 1); ni : ( 1; 1); mi : (1; 1); mi : ( 1; 1): (6.249)
In respect of the transformations (6.247) all the spin coeÆcients (6.236) can
be divided into two classes:
(a) quantities that are transformed in a uniform manner, e.g.,

 ! (Co )(C 1 _ )(Co )r_ (Co ) = C 3 C 1  ; (6.250)

(b) quantitites that are transformed in a nonuniform manner using deriva-
tives of C , e.g.,

 ! (Co )(C 1 _ )(C 1  )r_ (Co ) =

= CC  + C o _ r_ C:
1 1
(6.251)
If we take into account "primed" symmetry and spin and boost weights, the
main spinor quantities become
 : (3; 1);  : (3; 1);  : (1; 1);  : (1; 1);
0 : ( 3; 1);  0 : ( 3; 1); 0 : ( 1; 1);  0 : ( 1; 1);

0 = 04 : (4; 0); 1 = 03 : (2; 0);

2 = 02 : (0; 0); 3 = 01 : ( 2; 0);
4 = 00 : ( 4; 0); (6.252)
6.10. FORMALISM OF TWO-COMPONENT. . . 83

00 =  00 = 022 : (2; 2); 01 = 10 = 021 : (2; 0);

02 =  20 = 20 : (2; 2); 10 = 01 = 012 : (0; 2);
0
11 =  11 = 011 : (0; 0); 12 = 21 = 010 : (0; 2);
20 = 02 = 002 : ( 2; +2); 21 = 12 = 001 : ( 2; 0);
22 = 22 = 000 : ( 2; 2);
 =  = 0 = R=24 : (0; 0):
For weighted quantities we will introduce new dierential operators, such
that their action on a scalar  of the type fp; q g is dened as
P  = (D p" q"); P 0  = (
+ p"0 + q"0);
(6.253)
@ = (Æ p + q 0 ); @ 0  = (Æ + p q ):
Operators (6.253) have the following spin weights:
P : (1; 1); @ : (1; 1); (6.254)
P 0 : ( 1; 1); @ 0 : ( 1; 1):
In terms of (6.252) and dierential operators (6.253), we can write the spinor
equations (B s+) in a simpler form [53]. Shown schematically in Fig. 6.1 are
the boost and spin weights of the main spinors of the A4 geometry.
84 CHAPTER 6. GEOMETRY OF ABSOLUTE. . .

Figure 6.1: Boost and spin weights of main spinors of the A4 geometry
Chapter 7

Construction of solutions to
structural Cartan equations
of the geometry of absolute
parallelism
7.1 Selection of a frame of reference and
specialization of Newman-Penrose symbols
The structural Cartan equations of any geometry describe the general con-
nection between basic geometrical characteristics of a given geometry. A special
solution of structural equations determines specic geometrical quantities, such
as the curvature, connection, metric, etc., characteristic of a given specic solu-
tion [55]. For simplicity we will investigate the structural Cartan equations of
the A4 geometry
r[k eam] eb [k T ajbjm] = 0; (A)
Rabkm + 2r[k T ajbjm] + 2T ac[k T cjbjm] = 0; (B )
written in the vector basis, for their compatibility. These equations are essen-
tially a system that in the general case includes 44 (24 equations (A) and 20
equations (B )) nonlinear partial dierential equations of the rst order with the
following unknown functions:
(a) 6 components of anholonomic tetrad
ei a = raxi ; (7.1)
(b) 24 components of the Ricci rotation coeÆcients
T abk = ej b rk ea j ; (7.2)

85
86 CHAPTER 7. CONSTRUCTION OF SOLUTIONS. . .

(c) 20 components of the Riemann tensor

Ra bkm : (7.3)
Thus, in the general case we have 44 equations for 50 unknown functions.
This gives us some freedom in choosing a reference frame xi , of the tetrad ei a,
and also of the quantities T abk and Ra bkm . Therefore, search for specic solution
to the set of equations (A) and (B ) should rather be referred to as "constructing
solutions."
When constructing solutions it is convenient to represent the structural Car-
tan equations of the geometry of absolute parallelism in the spinor
-basis in
terms of Carmeli matrices
@C D_ Ai B_ @AB_ Ci D_ = (TC D_ )AP Pi B_ + Ai B_ (TDC R_ _
_ ) B
+

(TAB_ )C P Pi D_ Ci R_ (TBA R_ ;

_ ) D
+
_ (As)
RF ED
_ B_ = @DB
_ TF E
_ @F E_ TDB_ (TDB_ )F S TS B_ (TED F_
_ ) B_ TF F_ +
+

+(TF E_ )D S TS B_ + (TEF F_
_ ) B_ TDF_ + [TF E
+
_ ; TDB_ ]: (B s+ )
where the components of the traceless 2  2 matrices RF ED _ B_ and TF E
_ are found
from the relationships (6:88) and (6:103). Let us now nd the Newman-Penrose
symbols via the spinor representation of the invariant Haiashi derivative
AB_ i = rAB_ xi = @AB_ xi ; (7.4)
where the components of the spinor derivative @AB_ are denoted as
B_
_ _
@AB_ = A 0 1 (7.5)
0 D Æ
1 Æ

From the relationships (7.4)-(7.5) and
B_
0_ 1_
 iAB_ = A (7.6)
0 li= (Y ; V; Y ; Y )
0 2 3
mi = ( 0 ; !;  2;  3 )
m i = ( ; !;  ;  ) ni = (X 0; U;X 2; X 3)
0 2 3
1
we obtain
li = Dxi ; ni =
xi ; mi = Æxi ; mi = Æxi ; (7.7)
and also
Y 0 = Dx0 ; X 0 =
x0;  0 = Æx0 ;  = Æx0 ;
0

V = Dx1 ; U =
x1 ; ! = Æx1 ; ! = Æx1 ;

(7.8)
Y 2 = Dx2 ; X 2 =
x2;  2 = Æx2 ;  = Æx2 ;
2

Y 3 = Dx3 ; X 3 =
x3;  3 = Æx3 ;  3 = Æx3 :

7.1. SELECTION OF REFERENCE FRAME. . . 87

From the equality

@AB_ = AB_ iri = AB_ i ;i (7.9)
and the relationships (7.5) and (7.6) it follows
D = li ri;
= ni ri; Æ = mi ri; Æ = m iri (7.10)
or

D = V @ 1 + Y  @ ;
@x @x
@ @

= V 1 + X  ;
@x @x
Æ = w @1 +  @ ;
@x @x
@  @
Æ = w 1 + ; (7.11)
@x @x
 = 0; 2; 3:
Using these relationships, we write the vectors that make up the matrix (7.6)
as
li = V Æ1i + Y  Æi ;
ni = UÆ1i + X Æi ;
mi = !Æ1i +   Æi ;

mi = !Æ1i +  Æi : (7.12)
From the orthogonality condition for the Newman-Penrose symbols
iAB_ Aj B_ = Æij ; (7.13)

iAB_ Ci E_ = Æ A C Æ B_ E_ : (7.14)

follow the orthogonality conditions for the vectors (7.12)
li li = mi mi = m im i = ni ni = 0;
lini = mi m i = 1;
li m = li m = nimi = nim i = 0:
i i (7.15)
And from the formulas
gij = "AC "B_ D_ iAB_ jC D_ ; (7.16)
"00 = "11 = 0; "01 = 1; "10 = 1;
we nd
g ij = li nj + lj ni mi mj mj mi : (7.17)
Vectors that meet the orthogonality conditions (7.15) are null vectors and
in physics they are normally associated with propagation of radiation (i.e., with
88 CHAPTER 7. CONSTRUCTION OF SOLUTIONS. . .

matter that has no rest mass), where concepts of wave fronts, waves, rays, etc,
hold. In the process a family of null hypersurfaces u(xi ) = const is introduced.
We will take the vector li to be orthogonal to these hypersurfaces
li = u;i : (7.18)
Further, we will select the coordinates so that [60]
x0 = u;
x1 = r; where r is the aÆne parameter along the null
geodesics
x2 ; (7.19)
x3 ; ;
where x assign numbers to rays on each
2 3

hypersurface and are constant along the rays.

When selecting the coordinates, the vector li and li look like:
li = u;i = x0;i = Æi0; (7.20)
dxi dxi
li = 1 = = Æ1 i (7.21)
dx dr
or
li = (0; 1; 0; 0); li = (1; 0; 0; 0): (7.22)
From the orthogonality conditions
li ni = 1; limi = 0
it follows that
ni = (1; U; X 2; X 3);
mi = (0; !;  2;  3 ); (7.23)
and the relationships (7.11) become
D = V @1 = @ ;
@x @r
@ @  @ @ @  @

=U 1 + 0 +X = +U +X ;
@x @x @x @u @r @x
Æ = ! @1 +  @ = ! @ +  @ ;
@x @x @r @x
@  @ @  @
Æ=! 1 +  =! + ; (7.24)
@x @x @r @x
 = 2; 3:
Moreover, the vectors (7.12) will be given by
li = Æ1 i ;
ni = Æ0 i + UÆ1i + X  Æi ;
mi = !Æ1 i +   Æi ;

m i = !Æ1i +  Æi ; (7.25)
7.2. SPECIALIZATION OF THE SPINOR. . . 89

 = 2; 3:
Since
li = g ik lk = g ik Æk0 = g io = Æ1i ;
the metric tensor has the following structure [56]:
0 1
0 1 0 0
B 1 g 11 g 12 g 13 C
g ik = B
@ 0 g 12 g 22 g 23
C:
A (7.26)
0 g 13 g 23 g 33
Using the relationship (7.17) and (7.25), we get
g 22 = 2(U !! ); (a)
g 2 = X  (  ! +  ! ); (b) (7.27)
g Æ = (  Æ +   Æ ); (c)
; Æ;  = 2; 3:
As is seen from the above reasoning, the coordinates (7.19) selected and
the specialization of the Newman-Penrose symbols using the relationship (7.18)
made it possible to us to derive the dependence (7.27) and the general form
(7.26) of the metric tensor g ik of the A4 geometry.

7.2 Specialization of the spinor components of

the Ricci rotation coeÆcients
The spinor structural Cartan equations (As) and (B s) of the geometry A4 can
be viewed as a matrix of possible geometries of absolute parallelism that dier
in specic set of spinor geometrical characteristics. Therefore, we will assume
the solution of the spinor structural equations (As) and (B s ) (in this reference
frame) to be a set of variables consisting in the general case of:
(a) 6 independent components of the Newman-Penrose symbols
Ai B_ ; (7.28)
(b) 24 independent spinor components of the Ricci rotation coeÆcients
TAB_ ; TBA
_ ;
+
(7.29)
(c) 20 independent spinor components of the independent spinor components
of the Riemannian tensor
RABC
_ D_ ; R _
+
BADC_ ; (7.30)
that transform the equations (As) and (B s) into identities when substituted
into these equations.
90 CHAPTER 7. CONSTRUCTION OF SOLUTIONS. . .

In our search for solutions to the structural equations (As) and (B s) we will
rely on the symmetry conditions, and also on physical arguments, e.g., we will
subject the Riemannian tensor to the conditions of Einstein's vacuum
Rij = 0; (7.31)
which can be represented in terms of Carmeli matrices (7:103), (7:214) and
(7:217) as
RABC_ D _ = CABC _ D _ = 0:

We will now consider the limitations that can be imposed on the components
of the matrices (6:88), using physical reasoning. To this end, we will turn to the
relationship  
rk Ci D_ = (TAB_ )PC Pi D_ + Ci R_ (TBA
+ R_  AB_
_ )D_ k (7.32)
or
(TAB_ )PC Pi D_ + Ci R_ (TBA R_ = r  i  k :
_ )D (7.33)
+
_ k C D_ AB_
7.2. SPECIALIZATION OF THE SPINOR. . . 91

From (7.6), (6:88), and (7.33) we get

(7.34)
lk rk li = (" + " )li mi m i; (7:34a)
nk rk li = ( + )li mi mi ; (7:34b)
mk rk li = ( +  )li mi mi ; (7:34c)
m k rk li = ( +  )li mi mi ; (7:34d)
(7.35)
lk rk =ni (" + ")ni + m
i + m i; (7:35a)
n rk n = ( + )n + mi + m i;
k i i (7:35b)
mk rk ni = ( +  )ni + mi + mi ; (7:35c)
m k rk ni = ( +  )ni + mi + mi ; (7:35d)
(7.36)
lk r k mi = (" ")mi + l i ni ; (7:36a)
nk rk mi = ( )mi + li ni ; (7:36b)
mk rk mi = ( )mi + li ni; (7:36c)
mk rk mi = (  )mi + li ni ; (7:36d)
(7.37)
lk rk mi = (" ")mi + li ni; (7:37a)
nk rk m i = ( )m i + li ni ; (7:37b)
mk rk mi = (  )m i + li ni ; (7:37c)
m k rk mi = ( )mi + li ni: (7:37d)
Further, using the orthogonality condition (7.15), we have
 = rk li mi lk ;  = rk nim ink ;  = rk limi m k ;
 = rk nim m ;  = rk li mi mk ;  = rk ni mi mk ;
i k
 = rk li mi nk ;  = rk nim ilk ;
(7.38)
1
 = (rk lini mk rk mi mi m k );
2
 = 1 (rk li nimk rk mi mi mk );
2
1
= (rk li ni nk rk mi m ink );
2
92 CHAPTER 7. CONSTRUCTION OF SOLUTIONS. . .

" = 1 (rk lini nk

rk mi mi lk ):
2
On the other hand, we can write (7.32) as
(7.39)
rk lj = ( + )lk lj
lk mj lk mj + (" + ")nk lj nk mj
nk m j ( +  )mk lj + mk mj + (7:39a)
+mk mj ( +  )m k lj +
+mk mj + m k m j ;
rk nj = ( + )lk nj + lk mj +
+lk m j (" + ")nk nj + nk mj +
+nk m j + ( +  )mk nj (7:39b)
mk mj mk mj + ( +  )mk nj
m k mj m k mj ;
rk mj = ( )lk mj + lk lj
lk nj + (" " )nk mj + nk lj
nk nj + (  )m k mj (7:39c)
mk lj + mk nj + ( )mk mj
m k lj + m k nj ;
rk m j = ( )lk m j + lk lj lk nj +
+(" ")nk mj nk lj
nk nj + (  )mk mj (7:39d)
m k lj + mk nj + ( )mk m j
mk lj + mk nj ;
Alternating these relationships in the indices k and j gives
(7.40)
r k lj [ ] = 2<(")l[k nj ] (   )l[k mj ] (7:40a)
(   )l[k mj ] n[k mj ] n[k mj ] + 2i=()m[k m j ];

r k nj [ = 2<( )l[k nj ] (   )n[k mj ]

] (7:40b)
(   )n[k mj ] + l[k mj ] l[k m j ] + 2i=()m[k m j ];

r k mj
[ ] = ( +  )l[k nj ] + (2i=( ) + ) l[k mj ]+ (7:40c)
+l[k m j ] + (2i=(") ) n[k mj ] n[k mj ] (  )m[k m j ];

r k mj
[ ] = ( +  )l[k nj ] + ( 2i=( ) + ) l[k mj ]+ (7:40d)
+l[k m j ] + ( 2i=(") ) n[k m j ] n[k mj ] (  )m[k mj ] :
7.2. SPECIALIZATION OF THE SPINOR. . . 93

Convoluting the equations (7.39), we arrive at

(7.41)
rk ik = ( + ) + " + "; (7:41a)
rk nk = ( + ) +  + ; (7:41b)
rk mk =  +   + ; (7:41c)
rk mk =  +  +  + : (7:41d)
Relationships (7.34)-(7.41) appear to be quite useful for the specialization
of the spinor components of the Ricci rotation coeÆcients. Really, we will
require, for instance, that the isotropic vector li should obey the equations of
the geodesics of Einstein's gravitation theory
k k rk li = 0: (7.42)
It follows then from (7.34a) that
(" + ")li mi m i = 0: (7.43)
Clearly, the relationship (7.43) holds, if the spinor components of the Ricci
tensor have the following limitations:
(" + ") = 0;  =  = 0: (7.44)
The conditions of parallel transfer of the vectors mi , mi and ni along lk in
Einstein's gravitation theory become
lk rk mi = 0; lk rk m i = 0; lk rk ni = 0:
It follows from (7.35a), (7.36a) and (7.37a) that these relationships are valid
if
 =  =  =  = " = " = 0: (7.45)
The isotropic vector li is connected with the three optical parameters [40]:
(a) extension
1
 = ( + ) 2 = 0; 5rili ; (7.46)
(b) rotation
 1=2
1
! = ( )(2) 1 2= = r l rk li
2 [k i]
; (7.47)

(c) shift
 1=2
1
j^ j = (j j)1 2= = r l rk li
2 (k i)
 2
: (7.48)
94 CHAPTER 7. CONSTRUCTION OF SOLUTIONS. . .

Since we have taken li to be a gradient vector (li = u;i ), it follows from

(7.47) that
! = 0;  = : (7.49)
Furthermore, in that case we have [12]
 =  + : (7.50)

7.3 Specialization of the spinor components of

the Riemann tensor
It would be a good idea in our search for physically meaningful specic solu-
tions to the structural Cartan equations of the A4 geometry to use completely
geometrized Einstein's equations
2AB C_ D_ + "AB "C_ D_ = TACB
_ D_: (7.51)
It was shown in Chapter 6 that
1
 AC_ j  B D_ m (2AB C_ D_ + "AB "C_ D_ ) = Rjm g R;
2 jm
(7.52)

 AC_ j  B D_ m TACB
_ D_ = Tjm ; (7.53)
where the geometrized matter energy-momentum tensor Tjm is derived from
(7:221). Looking at various types of geometrized tensors (7:221), such as, e.g;
(a) energy-momentum tensor of the homogeneous A4 space

Tjm
(1)
= ~ gjm; 
~ = const; (7.54)
(b) Einstein's vacuum tensor
Tjm
(2)
= 0; (7.55)
(c) energy-momentum tensor of isotropic radiation
Tjm
(3)
= lj lm ; li li = 0 (7.56)
and so on, we will obtain various limitations to the spinor components of the
matrix RABC_ D_.
From the relationships (7.51)-(7.53) for tensors of the form (7.54) we will
nd the following limitations on the components of the matrices (6.103)
00 = 22 = 02 = 20 = 11 = 01 = 10 = 12 = 21 = 0;
0 6= 0; 1 6= 0; 2 6= 0; 3 6= 0; 4 6= 0; (7.57)
7.3. SPECIALIZATION OF THE SPINOR. . . 95

R
~ = = 6: (7.58)
4
At the same time, the condition (7.54) imposes, via (6.221), limitations on
the components of the matrices (6.88).
In the case of Einstein's vacuum the conditions (7.55) should be regarded
as equations to be satised by the components of the matrices (6.88). In the
process, in addition to (7.57), we will get
 = 0: (7.59)
For tensors of the form (7.56) we have
00 = 02 = 20 = 11 = 01 = 10 = 12 = 21 =  = 0;

 = 22 = ; (7.60)
2
0 6= 0; 1 6= 0; 2 6= 0; 3 6= 0; 4 6= 0:
To get an insight into the physical meaning of each spinor component of the
Weyl tensor 0 , 1 , 2 , 3 and 4 , we will consider ve cases:
(a) 0 6= 0, the other components are zero;
(b) 1 6= 0, same as above;
(c) 2 6= 0, - " -;
(d) 3 6= 0, - " -;
(e) 4 6= 0, - " -.
In each of the ve cases the components of the Weyl tensor have the following
algebraic properties according to Petrov:
(a) N type (or f4g) [57, 58] with the propagation vector ni;
(b) III type (or f31g) with the propagation vector ni;
(c) D type (or f22g) with the propagation vector li and ni ;
(d) III type (or f31g) with the propagation vector li;
(e) N type (or f4g) with the propagation vector li.
The propagation vector is meant to be the main light direction [40]. If in
the A4 space the condition of Einstein's vacuum Rjm = 0 is met, and the vector
li meets the equations
l[iRj ]km[nls]lk lm = 0; (7.61)
then the vector li corresponds to one of the four main light directions of the
Riemannian tensor, and we have
0 = 0: (7.62)
If two or more of the main light directions point along the propagation vector
li, then
Rijk[mln] lj lk = 0 (7.63)
or
0 = 1 = 0: (7.64)
96 CHAPTER 7. CONSTRUCTION OF SOLUTIONS. . .

According to the Goldberg-Sax theorem [59], it follows from (7.64) that

 =  = 0: (7.65)
A simple proof of the theorem is given in [40]. It relies on the second Bianchi
identities (D s+ ).
Similarly, it can be shown that from the condition
3 = 4 = 0 (7.66)
we get
 =  = 0: (7.67)

7.4 Construction of the asymptotic behavior of

insular-type geometries
A4 geometry is said to be an insular-type geometry, if at innity its main
characteristics (metric, connection, curvature) are identical to those of a at
space.
We will also assume that the conditions of Einstein's vacuum (7.31) and rela-
tionships (7.45), (7.49), (7.50) are valid. With these assumptions the structural
Cartan equations of the A4 geometry (As+), (B s+) and the second Bianchi
identities (D s+) can conveniently be split into the following three groups of
equations:

7.4.1 Radial equations containing derivatives with

respect to r


D  =   +  ; (7:68a)
D! = ! + ! ( +  );  (7:68b)
(7.68)
DX  = ( +  ) + ( +  ) ; (7:68c)
DU = ( +  )! + ( +  )! ( + ); (7:68d)
D = 2 + ; (7:69a)
D = 2 + 0 ; (7:69b)
D =  +  + 1 ; (7:69c)
D =  + ; (7:69d)
D =  +  + 1 ; (7:69e) (7.69)
D =  +  + 2 ; (7:69f )
D =  + ; (7:69g )
D =  +  + 2 ; (7:69h)
D =  +  + 3 ; (7:69i)
7.4. CONSTRUCTION OF THE ASYMPTOTIC. . . 97

D 1 Æ 0 = 4 1 4 0 ; (7:70a)
D 2 Æ = 3 2 2 1  0 ; (7:70b) (7.70)
D 3 Æ 2 = 2 3 2 1 ; (7:70c)
D 4 Æ 3 =  4 + 2 3 3 2 : (7:70d)

7.4.2 Nonradial equations


ÆX
  = ( + ) +   ; (7:71a)
Æ Æ  = ( )  + (  ) ; (7:71b) (7.71)
Æ! Æ! = ( )! + (  )! + ( ); (7:71c)
ÆU
! = ( + )! + ! ; (7:71d)

 Æ = 2 + ( 3  ) 4 ; (7:72a)
Æ Æ = ( + ) + ( 3) 1 ; (7:72b)
Æ Æ =   2 +  +  2 ; (7:72c)
Æ Æ = ( +  ) + ( 3 ) 3 ; (7:72d)
Æ
 =  +  + 2 2 + ; (7:72f ) (7.72)
Æ
 =  + ( + )  + ; (7:72g)
Æ
 = 2 + ( +  3 ) + ; (7:72h)

 Æ = ( + ) 2  2 ; (7:72i)

 Æ = ( ) +    3 : (7:72j )

7.4.3 U-derivative equations

0 Æ 1 = (4 ) 0 (4 + 2 ) 1 + 3 2 ; (7:73a)

1 Æ 2 =  0 + (2 2) 1 + 2 3 3 2 ; (7:73b) (7.73)

2 Æ 3 = 2 1 3 2 + ( 2 + 2 ) 3 +  4 ; (7:73c)

3 Æ 4 = 3 2 (2 + 4) 3 + (  + 4 ) 4 : (7:73d)
Suppose now that the structural Cartan equations of the A4 geometry de-
scribe an insular radiating system. In the process, the quantity 0 behaves at
an asymptotic along the coordinate r as
0 = o(r 5); (7.74)
whereas
D 0 = o(r 6): (7.75)
The conditions (7.74) has been chosen on purely physical grounds in such a
manner that the quadrupole radiation in a linear approximation of Einstein's
gravitational theory would correspond to the asymptotic. It is clear that we
could use another kind of asymptotic and have other asymptotic properties
98 CHAPTER 7. CONSTRUCTION OF SOLUTIONS. . .

for the insular A4 geometry. It is precisely what it mean by "constructing a

geometry."
Let now uniform perturbations in the coordinates x2 and x3 do not change
the nature of the asymptotic (7.74) and (7.75), i.e.,
d 0 = o(r 5); : : :; d d d dÆ 0 = o(r 5); (7.76)
d D 0 = o(r 6); : : :; d d d dÆ D 0 = o(r 6);
where
d = @  ; ; ; : : : = 2; 3:
@X
Using the relationships (7.74)-(7.76), we can nd the asymptotic behavior
of all the other spinor quantities that enter the equations (7.68)-(7.73). For
instance, we will show how the asymptotic behavior of the quantities  and  is
to be determined. We will introduce the matrices [40]
   
P=   ; Q= 0 0
  0: 0
Then the equations (7.69a), (7.69b) and their complex conjugates
D = 2 + ; D = 2 + ; (7.77)
D = 2 + 0 ; D = 2 + 0
will become
DP = P 2 + Q: (7.78)
This equation has a solution of the form
P = (DY ) 1 ; (7.79)
where  
Y = yy1 yy2 (7.80)
1 2

is a nonsingular solution (for a given P ) to the equations

DY = P Y: (7.81)
It is seen from (7.79) and (7.81) that the matrix (7.80) obeys the equations
D 2 Y = QY: (7.82)
The asymptotic behavior of the solutions of the equation (7.82) for the case
where Z
r j 0jdr = o(1);
has the form [40]
DY = F + o(1); (7.83)
7.4. CONSTRUCTION OF THE ASYMPTOTIC. . . 99

Y = rF + o(r ); (7.84)
where F is a constant matrix. Since in this case Q = o(r ), we will obtain
5

from (7.82) and (7.84)

D 2 Y = rQF + o(r 4 ) = o(r 4): (7.85)
Integrating this twice gives
DY = F + o(r 3 ); (7.86)

Y = rF + E + o(r 2); (7.87)

where E is a constant matrix. Solution (7.79) can now be written as
P = r 1 I + r 2 EF 1
+ o(r 3 ): (7.88)
Here E is a nonsingular matrix and I is a unit matrix. If F = 0, we have from
(7.88)
 = r 1 + o(r 2);  = o(r 2 ): (7.89)
Using other equations of the system (7.68)-(7.73) and working through the
same procedure we can nd for the quantities in these equations the following
asymptotic properties [40]:
  = o(r 1 ); ; ; ; ;  = o(r 1 );
X  ; ! = o(1); ; = o(1); (7.90)
U = o(r ); 1 = o(r 4);

2 = o(r 3 ); 3 = o(r 2); 4 = o(r 1):

To get a closer look at the asymptotic behavior of the quantities (7.89) and
(7.90) we will do the following [60]. We will write (7.89) as
 = r 1 + g (r); (7.91)
 = h(r );
where g; h = o(r 2 ).
Substituting (7.91) into (7.69a) and (7.69b) gives
Dg + 2r 1 g = g 2 + hh = o(r 4); (7.92)
Dh + 2r 1 h = 2gh + 0 = o(r 4):
We will at rst look for a solution to these equations to within the terms of
the order of magnitude o(r 3 ). Integrating the equations, we nd g (r ) [60]
R nR R o
g = e R 2dr=r e 2dr=r o(r 4 )dr + 0 = (7.93)

= r 2 o(r 4)dr + 0 = 0 r 2 + o(r 3 )
100 CHAPTER 7. CONSTRUCTION OF SOLUTIONS. . .

and a similar solution for h(r ). In (7.93) the sign 0 on the integration constant
 implies that the constant is independent of r . Hence
 = r 1 + 0 r 2 + o(r 3); 0 = 0 (u; x );
 =  0 r 2 + o(r 3 );  0 =  0 (u; x ); (7.94)

 = 2; 3:
Using the coordinate transformations r 0 = r r 0 (u; x ) we can eliminate the
term 0 =r 02, therefore
 = r 1 + o(r 3);
 =  0 r 2 + o(r 3):
Putting again
= r 1
+ g (r );  =  0 r 2
+ h(r );
where
g (r);h(r ) = o(r 3); (7.95)
and collecting all the terms in (7.69a) and (7.69b) up to those of the order of
magnitude o(r 5), we have
Dg + 2r 1 g = o(r 4) =  0  0 r 4 + o(r 5); (7.96)
Dh + 2r 1 h = o(r 5):
Integrating (7.95) gives
Z 
g=r 2
(  r
0 0 4
+ o(r ))r dr + C1 ;
5 2

Z 
h=r 2
r o(r ) + C2
2 5

or
g = C1 r 2  0  0 r 3 + o(r 4 );
h = C2 r 2 + o(r 4 ):
It follows from (7.95) that
C1 = C2 = 0;
therefore
 = r 1  0  0r 3 + o(r 4); (7.97)
 =  0 r 2 + o(r 4):
Going over the procedure, we can nd that
 = r 1  0  0r 3 + o(r 5); (7.98)
 =  0 r 2 + ( 0 0 0; 5 00 )r 4 + o(r 5):
7.4. CONSTRUCTION OF THE ASYMPTOTIC. . . 101

Similarly, we can nd the asymptotic behavior in r of other variables. The

following are the results obtained in [60]:
(a) for spinor components of the Riemannian tensor
(7.99)
0 = 00 r 5
+ o(r 6); (7:99a)

1 = 01 r + (40 00  00; )r 5 + o(r 6); (7:99b)
4 0

2 = 02 r 3
+ (20 01  0 01; )r 4 + o(r 5); (7:99c)
3 = 03 r 2
 0 02; r 3
+ o(r 4); (7:99d)
4 = 04 r 3
+ (20 03 +  0 03; )r 2
+ o(r 3); (7:99e)
 = 2; 3;
(b) for spinor components of the Ricci rotation coeÆcients
(7.100)
 = r 1  0  0 r 3 + o(r 5); (7:100a)
 =  0 r 2 + ( 0  0 0; 5 00 )r 4 + o(r 5 ); (7:100b)
 = 0 r 1 +  0 0 r 2 +  0  0 r 3 0 + o(r 4 ); (7:100c)
 = 0 r 1 +  0 0r 2 ( 0 0 0 + 0; 5 01 )r 3 + o(r 4); (7:100d)
 = 0; 5r 3 01 + 1 r 4 (20 00; 80 00 +  0 01 ) + o(r 5 ); (7:100e)
6
 = 0r 1  0 0 r 2 + (0; 5 01 +  0  0 0 )r 3 + o(r 4); (7:100f )
 = 0 r 1 ( 0 0 + 02 )r 2 + ( 0  0 0 0 01 +

+0; 5 01; )r 3 + o(r 4 ); (7:100g)
0

1 1 0 0
= 0 0; 5 02 r 2 + (  01;  1 0; 50 01 ) + o(r 4 ); (7:100h)
3 6

 = 0 03 r+ 0; 5 02; r 2 + o(r 3); (7:100i)
1 0

 = 2; 3;
(c) for the components of the Newman-Penrose symbols
(7.101)
1
U = ( 0 + 0 )r + U 0 0; 5( 02 + 02 )r 1
+ r 2 (0 01; +
6
+ 0 1; ) 2(0 01 + 01 ) + o(r 4); (7:101a)
0
102 CHAPTER 7. CONSTRUCTION OF SOLUTIONS. . .

X  = 1 r 3 ( 01 0 + 01  0) + o(r 4 ); (7:101b)
6
  =  0 r 1  0 0 r 2 +  0  0  0 r 3 + o(r 4); (7:101c)
! = ! 0 r 1 r 2 ( 0 ! 0 + 0; 5 01 ) + o(r 3 ); (7:101d)
 = 2; 3:
To simplify the remaining computations we will make use of the coordinate
transformations
r 0 = r + R0 (0; 2; 3) translations
0   (7.102)
u = u; x = x ; b = 2; 3; of the origin of r ,
r0 = r= ;_ u0 = (u) relabeling of
0 (7.103)
x = x ;
 hypersurfaces;
r0 = r; u0 = u;
relabeling of
0 (7.104)
x = x (0; 2; 3);
 geodesics:
From the equations (7.27a) and (7.27b) we have
  0 0
g  = (   +    ) = ( 0  +   0 )r 2
+ :::
;  = 2; 3: (7.105)
Using the coordinate transformations (7.102)-(7.104) we can reduce the met-
ric (7.105) to a conformally at metric [61,62]. Up to the terms of the order of
o(r 3 ), we have here
g 22 = g 33; g 23 = g 32 = 0: (7.106)
Since
g 22 = 2 02  02 r 2 + o(r 3);
g 23 = ( 02 03 +  02  03)r 2
+ o(r 3 );
g 33 = 2 03  03 r 2
+ o(r 3);
it follows from the conditions (7.106) that
 02 = i 03 = P (u; x ): (7.107)
The remaining coordinate transformations for the variables x2 and x3 look
like [62]
0 0
x2 + ix3 = f (x2 + ix3 ; u) (7.108)
We will next solve a set of nonradial equations (7.71) and (7.72) in order
to express the integration "constants" obtained in solving the radial equations
through only two functions 0 and P .
By way of example, we will consider the nonradial equation (7.72h)

 Æ = ( + ) 2  2 : (7.109)
7.4. CONSTRUCTION OF THE ASYMPTOTIC. . . 103

Using the denitions (7.24) of spinor derivatives, we will write (7.109) as

;0 + U;1 + X ; !;1
  ; ( + ) +
+2 +  + 2 = 0: (7.110)
Substituting here the necessary expressions from the solution (7.99)-(7.101)
and dierentiating with respect to r , we will equate the factors at various degrees
of 1=r to zero1. As a result, from (7:110), we will get:
(1) the factor at 1=r is identically zero;
(2) the factor at 1=r 2 is (U 0 0 ), whence
U 0 = 0 ;
(3) the factor at 1=r 3 is
( 0 0 );0 + 2 0  0 ( 0 + 0 ) ( 0 0 +  0  ) = 0;
0

and this relationship denes 0 and 0 , if the other terms are known;
(4) the factor at 1=r 4 is identically zero.
We will introduce the notation
r = @x@ @
@x32
+i (7.111)

then the nal expressions for the <<constants>> 0 ; 0 ;  0 ; : : : in terms of the two
main functions P and  become:
0 = 0; 5(ln P );0 ;
0 = 0; 5P r(ln P );0;
 = 0; 5P r(ln P P );0;

0

! 0 = P hr 0 2 0 r(ln Pi) ; (7.112)

=
0 =  0 ln( 0 P 1=2 P ) ;
3 2

;0
 = U = 0; 5P P rr ln(P P );
0 0

02 02 = (P r! 0 +
+20 ! 0 +  0  )
0

(P r!0 + 20! 0 +  0 0); (7.113)

03 = P r0 P r0 + 40  ;
0

4 = P r + 2 
0 0 0 0

0;0 4 0 0 :
The functions 02 + 02 , 00 and 01 in addition to the functions  0 and P
are the basis functions for insular-type systems.
1 For instance, if we have the asymptotic expression Ar 1 + Br 2 + Cr 3 + o(r 4 ) = 0
!1
(A; B; C are independent of r), then, multiplying this expression by r and putting r
!1
, we
will get A = 0. Further, multiplying by r2 and letting r , we will have B = 0; and so
forth.
104 CHAPTER 7. CONSTRUCTION OF SOLUTIONS. . .

Propagation of the functions 02 ; 00 and 01 in the u - direction is dened by

a group of equations (7.73). For instance, the equation (7.73a), to within terms
of the order of magnitude of o(r 6 ); becomes
00;0 r 5
+ 5( 0 + 0 ) 00 r 5
P r 01 r 5 0 00 r 5
20 01 r 5
3 0 02 r 5 + o(r 6) = 0;
hence
00;0 + 5( 0 + 0 ) 00
P r 01 4 0 00
20 01 3 0 02 = 0: (7.114)
The next two equations of group (7.73) (equations (7.73c) and (7.73d)) give
01;0 + 2( 0 + 2 0 ) 01 P r 02 2 0 03 = 0; (7.115)
02;0 + 3( 0 + 0 ) 02 P r 03 20 03 2 0 04 = 0: (7.116)
The last equation of group (7.73) is satised identically. Function P can be
chosen so that
P;0 = 0;
i.e.,
P = P (x2; x3 );  = 2; 3: (7.117)
Under this condition, the equations (7.112) and (7.113) are simplied dra-
matically to yield
0 = 0;
0 = 0; 5rP;
 = 0; (7.118)
! 0 = P 3r( 0=P 2 );
 =  ;0 ;
0 0

0 = P 2 rr ln P;
( 02 02 ) =
 
= P r(! 0 =P ) r(!0 =P ) +
2

+ 0 ;00  0  0;0 ;
03 = P r(P 2rr ln P )
P 3 ( 0;0 =P 2);
04 =  0;00: (7.119)
Equations (7.73) now become
00;0 r(P 01) 3 0 02 = 0;
01;0 P r 01 2 0 03 = 0;
(7.120)
02;0 P 2r( 03 =P )
 0  0;00 = 0:
7.4. CONSTRUCTION OF THE ASYMPTOTIC. . . 105

We can now write the nal form of the Riemannian metric. The metric of
the insular-type system looks like
0 1
0 1 0 0
B 1 g 11 g 12 g 13 C
g ik = B
@ 0 g 12 g 22 g 23
C;
A (7.121)
0 g 13 g 23 g 33
where
 
@ 2 + @ 2 ln P
g = 2P
11 2
( 02 + 02 )r 1
+
@x22 @x32
" ( 0
)  
1 1 01
+ P2
3
r P +r
P
   
 r 
0 0
6P 4 r r 2
+ o(r 3 );
P2 P2
g 12 = r 2 <(f ) + r 3 <(h) + o(r 4);
g 13 = r 2 =(f ) + r 3 =(h) + o(r 4);
P = P (x2 ; x3 ); (7.122)
f = 2P 4r( 0 =P 2); 
h = 4P 13 01 + P 3 0 r( 0 =P 2) ;

g 22 = 2P 2 r + 2P ( 0 +  0 )r 3
2

6 0  0 P 2r 4 + o(r 5 );
g 23 = 2iP 2 ( 0  0 )r 3 + o(r 5 );
g 33 = 2P 2 r 2 2P ( 0 +  0 )r 3
6 0  0 P 2r 4 + o(r 5):
In matrix (7.121) the component g 11 can be worked out to within the terms
of the order of magnitude of o(r 4), and the terms g  (;  = 2; 3) to within
o(r 5).
If now we specify the initial conditions 02 + 02 , 00 , 01 ,  0 and P at innity,
then the problem of the initial values will be overcome.
The zero surface of initial values u0 is determined by the condition

!1 ( 0 r ) < 1:
00 = rlim 5

The initial value  0 is dened on the world tube at spatial innity. On the
tube we chose
 0 = rlim
!1 (r )
2

as an independent function of the variables u, x2 ; and x3 . The remaining initial

data are taken on a two-dimensional surface at innity that is determined by the
106 CHAPTER 7. CONSTRUCTION OF SOLUTIONS. . .

intersection of the zero surface u0 and the world tube. On that two-dimensional
surface we specify

!1 ( 1 r ); 2 + 2 = rlim
!1 r ( 2 + 2 )
01 = rlim 4 0 0 3

and
P 2Æ  = rlim  2
!1 (g r )
as functions of x2 and x3 .

7.5 Classication of solutions to the structural

Cartan equations of the A4 geometry by isom-
etry groups
It was shown in Chapter 6 that the structural Cartan equations of the A4
geometry can be taken to be gauge equations with gauge groups T4 and O (3:1).
Knowledge of this fact alone is not suÆcient for one to be able to tell vari-
ous specic solutions of the structural Cartan equations from one another by
group behavior. This is made possible by the technique of the embedding of A4
geometries into a at space Ep of many dimensions (N > 4).
We will make the following assumption:
(1) we will take the A4 space to be a continuous deformation of the Minkowski
space E4(3.1);
(2) we will suppose that A4 has a minimal at embedding space Ep (r; s)
of dimensionality p = r + s, where the signature r + s means r positive and s
negative diagonal elements of the metric tensor  (;  = 1; 2; : : :; p) of space
Ep (r; s).
Let now X  be Cartesian coordinates of the at space Ep(r; s), and x be
Gaussian coordinates based on A4 that is embedded into Ep (r; s). Here and
later in the section the Greek indices assume the values 1; : : :; N .
We will denote the coordinates of a point in space A4 by xi ; the coordinates
in directions orthogonal to A4 , by xA . Here and in the section the small-cap
Roman indices i; j; k; : : : assume the values 1,2,3,4, and the large-cap Roman
indices A; B; C : : : the values 5 : : : N:
In the embedding the coordinates are transformed as follows:
X  = X (x ); (7.123)
and the tensors between these two reference frames are transformed using the
derivatives
@x ; x = @x ;
x = @X   @X
 (7.124)
X = @x ; X = @X
 @X 
@x :
7.5. CLASSIFICATION OF SOLUTIONS. . . 107

Thus, if  are Cartesian components of the metric tensor of a space

Ep (r; s), then its Gaussian components are
g  = x x   ; g = XX  : (7.125)

The inverse relationship are


  = XX g   = x x g (7.126)
If the A4 space has an isometry group, this group consists of pseudorotations
and re ections O (r; s) of the space Ep(r; s).
Suppose now that we have a coordinate transformation in Cartesian coordi-
nates 0
X  = X  + U  ; U  = "X  (7.127)
that is essentially one innitesimal transformation of the group O (r; s), such that
N (N 1)=2 of innitesimal quantities e  are constant and meet the condition
e( ) = 0. The isometric nature of the transformation (7.127) is manifested by
the fact that the Lee derivative with respect to U  of   vanishes
L  = U ; ( )
= 0: (7.128)
On the other hand, the Gaussian coordinates are transformed as
0
x = x +   ; (7.129)
where   = x U  are group generators.
Relationship (7.129) can be split into two parts
xi0 = xi +  i; xA0 = xA +  A ; (7.130)
i = 1; 2; 3; 4; A = 5 : : :p:
The embedded A4 space in a Gaussian reference frame is now subject to the
condition
xA = 0: (7.131)
If f (x ) is any real function dened in Ep (r; s), then its space-time part will
be
f (x ) = f (x )jA4 at xA ! 0:
A function dened only on an embedded A4 surface will be denoted as f (A4 ).
For instance, we will have
 j 
g ij A4 = xi x    = g ij A4 :
A4
The Killing equations for the vector   look like

Lg  =r (  ) = 0 (7.132)
108 CHAPTER 7. CONSTRUCTION OF SOLUTIONS. . .

where   
r j i
( )
= 0; r (A  i) = 0; r (  A) = 0: (7.133)
To the transformations (7.130) in an embedded A4 correspond the transfor-
mations 0 0
xi = xi +  i jA4 = 0; xA = xA +  A jA4 = 0 (7.134)
on which the following conditions are imposed:
  
r (j  i) jA4 = 0; r (A  i) jA4 = 0; r (  A) jA4 = 0: (7.135)
The covariant derivative of the vector  in a Gaussian reference frame with

respect to the connection
 is

r  = ; +
  = 0: (7.136)

It is seen that the expression for r  does not coincide with the expression
for the covariant derivative in A4 space, unless the following condition is met:
xA jA4 = 0: (7.137)
Condition (7.137) has the meaning that transformation (7.129) does not
change the denition A4 . Condition (7.137) identies in the group O (r; s) a
subgroup that denes the symmetry of the embedded A4 space. By adding
the group O (r; s) re ections and the condition (7.137), we will get the isometry
group of A4 space. Since the maximal dimensionality of an embedded space for
Riemannian spaces of dimensionality 4 is 10, then going over the signatures of
embedded spaces makes it possible to establish 22 isometric groups [63].
Given in table 7.1 are Lee isometric groups for various specic A4 spaces
and their spinor representations.
The table also provides the most important subgroups. It is suÆcient to
specify one of the groups in the table to give an isometric denition of the
appropriate A4 geometry. On the other hand, each solution of the structural
Cartan equations of the A4 geometry has corresponding to it an embedded
space.
Shown in table 7.2 are some minimum embedded spaces for a series of A4
spaces that feature various Riemannian metrics [63].
All these spaces can be derived as solutions to the structural Cartan equa-
tions of A4 geometry (e.g., Riemannian metric of Godel space has been obtained
in Ozsvath [50] using the Newman-Penrose method, i.e., as a solution of the
structural equations).

7.6 A4 geometry with a Schwarzschild-type met-

ric
In order to construction an A4 geometry that has a Schwarzschild metric the
following condition must be met: (7.45), (7.49), (7.50), (7.55), (7.59), (7.65),
7.6. GEOMETRY A4 . . . 109

Table 7.1:

p Ep (r; s) Lp (r; s) Spinor Most important

group subgroups
4 E4 (3.1) SO (3:1) SL(2:C )
4 E4 (2:2) O (3:1) SU (1:1)  SU (1:1)
5 E5 (4:1) SO (4:1) SL(4:C ) SU (2)  SU (2)
5 E5 (3:2) SO (3:2) SU (1:1:1:1) SU (1:1)  SU (1:1)
6 E6 (5:1) O (5:1) SL(4:C )
6 E6 (4:2) O (4:2) SU (2:2) SU (2)  SU (2)
6 E6 (3:3) O (3:3) SL(4:C ) SU (1:1)  SU (1:1)
7 E7 (6:1) SO (6:1) SL(8:C ) SU (4)
7 E7 (5:2) SO (5:2) SU (2:2:2:2) SU (2:2)
7 E7 (4:3) SO (4:3) SL(8:C ) SU (2)  SU (2)
8 E8 (7:1) O (7:1) SL(8:C ) SU (4)
8 E8 (6:2) O (6:2) SU (1:1)  SU (4:4) SU (4)
8 E8 (5:3) O (5:3) SL(16:C ) SU (2)  SU (2)
8 E8 (4:4) O (4:4) SU (1:1)  SU (2:2:2:2) SU (2)  SU (2)
9 E9 (8:1) SO (8:1) SL(16:C ) SU (4)
9 E9 (7:2) SO (7:2) SU (4:4:4:4) SU (4:4)
9 E9 (6:3) SO (6:3) SL(16:C ) SU (4)
9 E9 (5:4) SO (5:4) SU (2:2:2:2:2:2:2:2) SU (2)  SU (2)
10 E10 (9:1) O (9:1) SL(16:C )
10 E10 (8:2) O (8:2) SU (8:8) SU (8)
10 E10 (7:3) O (7:3) SL(16:C )
10 E10 (6:4) O (6:4) SU (4:4:4:4) SU (4)
10 E10 (5:5) O (5:5) SL(16:C )

Table 7.2:
Ep(r:s) Metric of immersed space
E (4.1) De Sitter-Einstein space
E6 (5.1) Kruskal space
E6 (4.2) Schwarzschild space
E7 (5.2) Petrov space T2=C 4=4 [22]
E7 (4.3) Petrov space T1=C 4=5; 6
E9 (6.3) Robinson-Trautman space C  0
E9 (5.4) Robinson-Trautman space C  0
E10 (6.4) Axial-symmetrical Weyl space
E10 (5.5) Godel space
110 CHAPTER 7. CONSTRUCTION OF SOLUTIONS. . .

(7.66) and (7.67). The physical meaning of these constraints has been considered
earlier in the book. As a result the structural Cartan equations (7.69)-(7.73)
become:
(1) radial equations that contain a derivative with respect to r
D  =   ; (7:138a)
D! = ! ( +  ); (7:138b)
(7.138)
DX  = ( +  ) + ( +  ) ; (7:138c)
DU = ( +  )! + ( +  )! ( + ); (7:138d)
D = 2 ; (7:139a)
0 = 0; (7:139b)
D = ; (7:139c)
D = ; (7:139d)
D = ; (7:139e) (7.139)
D =  +  + 2 ; (7:139f )
0 = 0; (7:139g)
D =  + 2 ; (7:139h)
0 = 0; (7:139i)
0 = 0; (7:140a)
D 2 = 3 2 ; (7:140b) (7.140)
Æ 2 = 0; (7:140c)
0 = 0; (7:140d)
(2) nonradial equations
ÆX 
  = ( + )  ; (7:141a)
Æ Æ  = ( )  + (  )  ; (7:141b) (7.141)
Æ! Æ! = ( )! + (  )! + ( ); (7:141t)
ÆU
! = ( + )!; (7:141c)
0 = 0; (7:142a)
Æ = ( + ); (7:142b)
Æ Æ =  2 +  +  2 ; (7:142c)
Æ = ( +  ); (7:142d)

 =  +  + 2 ; (7:142e) (7.142)
Æ
 =  + ( + ); (7:142f )
Æ = 2; (7:142g )

 Æ = ( + ) 2 2 ; (7:142h)

 Æ = ( ); (7:142i)
(3) U -derivative equations
0 = 0; (7:143a)
Æ 2 = 3 2 ; (7:143b) (7.143)

2 = 3 2 ; (7:143c)
0 = 0: (7:143d)
7.6. GEOMETRY A4 . . . 111

In the reference frame we have chosen the structural equations have the form
@ @ @
; Æ = ! +   ;
D=
@r @r @x
@ @ @ @  @

= + U + X  ; Æ = ! +  ;
@u @r @x @r @x
 = 2; 3:
and the commutation relation
r arb
[ ] =
::c
ab (7.144)
become
(7.145)

D D
= ( + )D ( +  )Æ ( +  )Æ; (7:145a)
ÆD DÆ = ( +   )D Æ; (7:145b)
Æ

Æ = (   )
+ ( + )Æ; (7:145c)
ÆÆ ÆÆ = ( + )D + ( )
( )Æ (  )Æ; (7:145d)
therefore we can proceed with integration of the equations. Integration be-
gins with the radial equations that contain the derivative D. For instance, the
solution to the equation (7.139a), with the condition  =  has the form
 = r 1: (7.146)
Dierentiating (7.139a) with respect to Æ gives
ÆD = 2Æ: (7.147)
Applying the complex-conjugate operator (7.145b) to , we have ( = 0)
(ÆD DÆ ) = ( +  )D Æ; (7.148)
whend by (7.147), we have
DÆ 3Æ = 2 ( +  ): (7.149)
Using (7.139a), (7.139d) and (7.139e), we obtain the general solution (7.149)
Æ = ( +  ) 2 0 3 ; (7.150)
where  0 is an integration constant. Calculating (
D D
) , (Æ

Æ )
and (ÆÆ ÆÆ) (here and later we will use the notation 2 = ; 02 = 0 )
using the relationships (7.140b,c), (7.143b,c) and (7.145), we will arrive at three
new equalities

 + D = ( + ) ; (7.151)
Æ = ( +  ); (7.152)
112 CHAPTER 7. CONSTRUCTION OF SOLUTIONS. . .

Æ +
 = ( +  ) +  ( ): (7.153)
Substituting (7.150) into (7.152) and integrating, we will get   = 0, 0 2

whence  0 =  0 = 0.
Integrating the remaining radial equations gives [65]
 =  0 ;
 = 0 + 2 ;
 =  0 ;
= 0 +  0 0 +  0  0 + 2 1=2 0 ;
= 3 0 ; (7.154)
 n o
U = U 0 r ( 0 + 0) +   0! 0 1=2 +  0!0 1=2 ;
0 0

! = ! 0 + 0 +  0 ;
  =  0 ;
0
X  = X 0 +  0  0 +  0  ;  = 0; 2; 3:
Applying the operator (7.145a) to  and using the equation (7.139a), we will
obtain
2
 D
 = 2 ( + ) Æ Æ: (7.155)
Solving this equation using (7.142b), (7.150), (7.154) and considering that
 0 = 0, we get

 = M 0 2 +  0(0 +  0 )2 + ( 0 + 0 ) +
 
+ 0 (0 +  0 )  0 0 2 1=23 ( 0 + 0 ): (7.156)
Substituting this relation into (7.151) and integrating gives
 = 0 + M 0 + 1=22 ( 0 + 0 ): (7.157)
The next phase of integration consists in substituting the derived solutions
of the radial equations (7.146), (7.154) and (7.155) into the remaining unused
equations. After dierentiating with respect to r , we will equate to zero the
factors at the same degrees of 1=r . We will end up with a set of equations for
quantities independent of r .
Applying the operators Æ; Æ and
to  = 1=r , we have
Æ = !2 ; Æ = !2 ;
 = U2 : (7.158)
Comparing these equalities with (7.142b), (7.150) and (7.156) gives
M 0 = M 0; (7.159)
! 0 = 0; (7.160)
7.6. GEOMETRY A4 . . . 113

U 0 =  0(0 +  0 ) +  0 (0 +  0 )  0 0 M 0: (7.161)
From (7.142c) we get
0 = 0; (7.162)

 ;0 =  0 (0 + 3  0 ) + 1=2( 0 0 );
0 0
(7.163)
0 ; = 2  0 :
0
0
(7.164)
Consequently, from the equations (7.142g), (7.142d) and (7.142c) we will
have
 0 ;0 =  0 (30 +  0 ); (7.165)
 0 ;
0
=  0 (20  0 );
 0 0;  0 ; = 2 0 ( 0 ) + M 0 ;
0
0

0 M;0 = 2M 0 (0 +  0 ); (7.166)
 0 = 0:
Substitution of the last of these equations into the equality (7.161) gives
U0 = M0: (7.167)
From (7.142f-i) and (7.153) we get

X 0 ;0 =  0 ( 0 + 3 0 );
 0 M;0 = 2M 0 (0 +  0 );
X 00; 0 ; = 0 (0  );
0 0

X 0;
0
 0 ;
0
= 0 ( 0 + 0) 2 0  0 ;
X 0 M;0 = 2M 0 (0 + 0 ):
Equations (7.141) enables us to write
 0 X;0 X 0 ;
0
= 2 0  0 (0 +  0 )X 0 ;

 0 ;
0
 0  0; = 2 0  0 + 2 0 0 ;
;  = 0; 2; 3:
Substituting = 3 0 into the equations (7.140c), (7.143b) and (7.143c),
we will obtain
 0 0; = 3 0 (0 +  0  0 ); (7.168)

 0; = 3 0 (0 +  ):
0 0
(7.169)
X 0 0; = 3 0 ( 0 + 0 + 0 ); (7.170)
114 CHAPTER 7. CONSTRUCTION OF SOLUTIONS. . .

Let us consider the case of

0 = ; 0 = const:
0
(7.171)
Then from (7.166) and (7.167) and also from (7.164) and (7.160)) it follows that

0 + 0 = 0; 0 +  0 = 0: (7.172)
We will constrain ourselves to the case  0 = 0: Considering that  0 = 0, we
will obtain that  = 0: Equations (7.141) will then become
(7.173)
X 0M;0 = 0; (7:173a)
 0 M;0 = 0; (7:173b)
X 0 0;  0 ;
0
= 2 0 0; (7:173c)
0 0
 0 0;  ; = 400 + M 0 ; (7:173d)
 0 X;0 X 0;
0
= 2 0  0 ; (7:173e)
 0 ;
0
 0 0; = 2 0  0 + 2 0  0 : (7:173f )
Next we carry out the transformations
0
x = x (x ):
We will thus have the relation X 0 = Æ  0 satised. Now the only arbitrary
element in the selection of coordinates is the transformations
(7.174)
0
x0 = x0 + f (x2 ; x3 ); (7:174a)
20
x = g (x2; x3 ); (7:174b)
30
x = h(x2 ; x3 ): (7:174c)
Integrating (7.173) gives
M 0 = const:
Using the transformations
0 0 0
li = li ; ni = ni; mi = mi exp[i 0(x )]; (7.175)
we can achieve
0 = 0: (7.176)
It follows from this and (7.173e) that  0 is independent of x0 .
7.6. GEOMETRY A4 . . . 115

We will introduce the notation

 02 = P (x2 ; x3 );  03 = iP (x2 ; x3 ) (7.177)
and use the transformations (7.175) to make P a real function. The only arbi-
trariness will be the transformation of the coordinates and the components of
the light tetrad of the form
li0 = [A0 (x )] 1 li ; ni0 = A0 (x )ni; mi0 = mi ; (7.178)
r 0 = A0 (x )r
with the constant A0 in the transformation (7.174a) and the transformation
 = g ( );
where
 = x2 + ix3 : (7.179)
Using the notation (7.111), we will write the equations (7.173d) and (7.173f)
in the form
" #  
00 = r  00 ;
 = 1=2rP; r
0
P P
p 1
"0 = M 0 = 2P 2rr ln( 2P ) 2 :
Using the remaining arbitrariness and the selection of the coordinates and
components of the light tetrad enables us to write the solutions of these equa-
tions as
p
2P = 1 + 1=2"0 ;  00 = 0; 0 = 1=2"0 ;
"0 = 1=2; 0:
We now use the results obtained for the components of the Newman-Penrose
symbols to obtain
0i 0_ = (0; 1; 0; 0); 1i 1_ = (1; U; 0; 0); 0i 1_ = (0; 0; P; iP );

1
i00_ = (1; 0; 0; 0); i11_ = ( U; 1; 0; 0); i01_ = 2P (0; 0; 1; i):
Using the relationship
gij = "AC "B_ D_  ABi_  C Dj_
we can now derive the metric tensor gik
0 1
2U 1 0 0
B 1 0 0 0 C
gik = B
@ 0 0 (22 P 2 ) 1 0
C;
A (7.180)
0 0 0 (2 P )
2 2 1
116 CHAPTER 7. CONSTRUCTION OF SOLUTIONS. . .

where
U = "0 + 0 =r: (7.181)
(7.182)
Let now e0 = 1=2, it is then convenient to go over to the coordinates
Z
ct = x 0
dr=2U; r = x1 ;

( )1=2 x3
sin  = ; tg' = x2 :
(1 + 1=4 )
We will end up with the Riemannian metric
    1
2 0 2 0
ds = 1
2
c dt
2 2
1 dr 2 (7.183)
r r
r 2 (d 2 + sin2 d'2 );
that coincides with the metric of the Schwarzschild space at
0 = MG=c2 : (7.184)
Notice that, unlike the Schwarzschild metric of Einstein's theory, the met-
ric (7.183) is dened on a translations group T4 of the geometry of absolute
parallelism.
At 0 = 0 and 0 = 1=2 we have two more solutions that describe spheri-
cally symmetrical objects with mass M (not necessary rest mass), which move
at light and faster-than-light velocities
   
2 0 2 0 1

ds =2
r c dt
2 2
r dr 2 (7.185)

r 2 (d 2 +  2 d'2 );
   
2 0 2 0 1

ds =
2
1 c dt
2 2
1 dr 2 (7.186)
r r
r 2 (d 2 + sh2 d'2 ):
Combining all the results, we will write

Main geometrical characteristics of A4 geometry with a Rieman-

nian metric of Schwarzschildian type
(7.187)
1. Coordinate u; r; x2 and x3 are given by (7.19).
2. Components of the Newman-Penrose symbols
7.7. SOME PHYSICALLY MEANINGFUL. . . 117

0i 0_ = (0; 1; 0; 0); 1i 1_ = (1; U; 0; 0); 0i 1_ = (0; 0; P; iP );

1
i00_ = (1; 0; 0; 0); i11_ = ( U; 1; 0; 0); i01_ = (0; 0; 1; i);
2P
U = 1=2 + 0 =r; P = (2) =
1 2
(1 + =4);  = x2 + ix3 ;
0 = const:
3. Spinor components of the Ricci rotation coeÆcients
 = 1=r;  =  = 0 =r; = 0 =2r;
 = "0 =r + 2 0 =r 2;  = =4:
4. Spinor components of the Riemannian tensor

= 0 =r 3 :
Substituting the components of the Ricci rotation coeÆcients of the solution
(7.187) into the rotational Killing-Cartan metric, we obtain

r ) d 2
d 2 = ( 4) dx20
0 2
2( 0
(7.188)
2r r
2( 0 r ) sin2  2
d' :
r

7.7 Some physically meaningful solutions of the

structural Cartan equations of A4 geometry
Skipping detailed computations, we will simply provide some exact solutions
of the structural Cartan equations of the A4 geometry, which are given physical
interpretation in the theory of physical vacuum.

7.7.1 Solution with a variable source function

(7.189)
1. Coordinates x0 = u; x1 = r; x2 = ; x3 = '.
2. Components of the Newman-Penrose symbols
0i 0_ = (0; 1; 0; 0); 1i 1_ = (1; U; 0; 0); 0i 1_ = (0; 0; P; iP );
118 CHAPTER 7. CONSTRUCTION OF SOLUTIONS. . .

1
i00_ = (1; 0; 0; 0); i11_ = ( U; 1; 0; 0); i01_ =
(0; 0; 1; i);
2P
U (u) = 1=2 + 0 (u)=r; P = (2) 1=2 (1 + =4);  = x2 + ix3 ;
0 = 0 (u):
3. Spinor components of the Ricci rotation coeÆcients
 = 1=r;  =  = 0 =r; = 0 (u)=2r 2;
 = 1=2r + 0 (u)=r 2; 0 = =4:
4. Spinor components of the Riemannian tensor
@ 0 1
2 = = 0 (u)=r 3; 22 =  = _ 0 (u)=r 2 = :
2 @u r
The Riemann metric of the solution (7.189) in the coordinates (7.182) has
the form
   
2 0 (t) 2 2 2 0 (t) 1 2
ds2 = 1 c dt 1 dr (7.190)
r r
r 2 (d 2 + sin2 d'2 ):

7.7.2 Solution with quark interaction

(7.191)
1. Coordinates x = u; x = r; x = ; x = '.
0 1 2 3

2. Components of the Newman-Penrose symbols

0i 0_ = (0; 1; 0; 0); 1i 1_ = (1; U; 0; 0); 0i 1_ = (0; 0; P; iP );
i00_ = (1; 0; 0; 0); i11_ = ( U; 1; 0; 0); i01_ = 1 (0; 0; 1; i)
2P
~ 1=2
U = 1=2 + r ; P = (2) (1 + =4);  = x2 + ix3 :
2

3. Spinor components of the Ricci rotation coeÆcients

 = 1=r;  =  = 0 =r; = ~ r;
 = 1=2r ~ r 2 ; 0 = =4:
4. Spinor components of the Riemannian tensor
 = ~ =6 = R=24 = const:
The metric of the Riemannian solution (7.191) in the coordinates (7.182)
has the form



ds2 = 1 r 2=3 c2dt2 1 r 2 =3 1 dr 2 (7.192)
r 2 (d 2 + sin2 d'2 ):
7.7. SOME PHYSICALLY MEANINGFUL. . . 119

7.7.3 Solution with a short-range (nuclear) interaction

(7.193)
1. Coordinates x0 = u; x1 = r; x2 = ; x3 = '.
2. Components of the Newman-Penrose symbols
p
0i 0_ = (0; 1; 0; 0); 1i 1_ = (1; U; 0; 0); 0i 1_ =  ( irN = 2; 0; P; iP ); 1i 0_ = 0i 1_
p p 1
i00_ = (1; 0 rN x3 = 2 P; rN x2 = 2P ); i01_ = 2P (0; 0; 1; i); i10_ = i01_ ;
p p
i11_ = ( U; 1; UrN x3 = 2 P; UrN x2 = 2P ):
U = 1 + rN2 ; P = (2) =
1 2
(1 + =4);  = x2 + ix3 ;
2
rN = const:
3. Spinor components of the Ricci rotation coeÆcients
 = (r + irN ) 1 ;  = 0 ;  = ; 0 = =4;
= 2 0 =2;  = =2 + 2 0 =2 +  0 =2; 0 = irN :
4. Spinor components of the Riemannian tensor
2 = = 0 3 :
The metric of the Riemannian solution (7.193) in the coordinates (7.182)
has the form
ds2 = [cdt + 4rN sin2 (=2)d']2 + dr 2 = (7.194)
(r 2 + rN2 )(d 2 + sin2 d');
where
2rN2
=1 : (7.195)
r + rN2
2

7.7.4 Solution with an electronuclear interaction

(7.196)
1. Coordinates x = u; x = r; x = ; x = '.
0 1 2 3

2. Components of the Newman-Penrose symbols

0i 0_ = (0; 1; 0; 0); 1i 1_ = (1; U; 0; 0);
p
0i 1_ =  ( irN = 2; 0; P;iP ); 1i 0_ = 0i 1_
120 CHAPTER 7. CONSTRUCTION OF SOLUTIONS. . .

p p
i00_ = (1; 0 rN x3 = 2 P; rN x2 = 2P );
p p
i11_ = ( U; 1; UrN x3 = 2 P; UrN x2 = 2P );
1
i01_ = (0; 0; 1; i); i10_ = i01_ ;
2P
U = 1 +  (rre=2 + rN2 ); P = (2) =
1 2
(1 + =4);
2
 = x2 + ix3 ; rN = const; re = const:
3. Spinor components of the Ricci rotation coeÆcients
 = (r + irN ) 1 ;  = 0 ;  = ; 0 = =4;
= 2 0 =2;  = =2 + 2 0 =2 +  0 =2; 0 = re =2 + irN :
4. Spinor components of the Riemannian tensor
2 = = 0 3 :
The Riemannian metric of the solution (7.196) in the coordinates (7.182)
has the form
ds2 = [cdt + 4rN sin2 (=2)d']2 + dr 2 = (7.197)
(r + rN )(d + sin d');
2 2 2 2

where
rre + 2rN2
=1 : (7.198)
r 2 + rN2

7.7.5 Solution with electronuclearquark interaction

(7.199)
1. Coordinates x = u; x = r; x = ; x = '.
0 1 2 3

2. Components of the Newman-Penrose symbols

0i 0_ = (0; 1; 0; 0); 1i 1_ = (1; U; 0; 0);
p
0i 1_ = ( irN = 2; 0; P; iP ); 1i 0_ = 0i 1_
p p
i00_ = (1; 0 rN x3 = 2 P; rN x2 = 2P );
p p
i11_ = ( U; 1; UrN x3 = 2 P; UrN x2 = 2P );
1
i01_ = (0; 0; 1; i); i10_ = i01_ ;
2P
1
U= + (rer=2 + rN2 8~ rN4 ) + (
~ r 2 + 5rN2 );
2
7.7. SOME PHYSICALLY MEANINGFUL. . . 121

 = x2 + ix3 ; P = (2) 1=2 (1 + =4);

rN = const; re = const; ~ = const:
3. Spinor components of the Ricci rotation coeÆcients
 = (r + irN ) 1 ;  = 0 ;  = ; 0 = =4;
= 0 + 2 0 =2 ~ r;  = =2 + 2 0 =2 +  0 =2 ~ r 2 ;
0 = re=2 + irN = const; 0 = i~ rN :
4. Spinor components of the Riemann tensor
2 = = 0 3 ;  = ~ =6 = R=24 = const:
The metric of the Riemann solution (7.199) in the coordinates (7.182) has
the form
ds2 = [cdt + 4rN sin2 (=2)d']2 + dr 2 = (7.200)
(r 2 + rN2 )(d 2 + sin2 d');
where
rre + 2rN2 16~ rN4 ~ r 2 + 5rN ):
=1 2( 2
(7.201)
r 2 + rN2

7.7.6 Solution with Coulomb-Newton interaction and three-

dimensional rotation of a source

(7.202)
1. Coordinates x0 = u; x1 = r; x2 = ; x3 = '.
2. Components of the Newman-Penrose symbols
0i 0_ = (0; 1; 0; 0); 1i 1_ =  (
; Y; 0; a);

0i 1_ = p (ia sin ; 0; 1; icosec ); 1i 0_ = 0i 1_ ;

2
i00_ = (1; 0; 0; a sin2  );


i11_ =  Y; () 1 ; 0; a sin2 Y ;


i01_ = p ia sin ; 0; () 1 ; i
sin  ; i10_ = i01_ ;
2

= r 2 + a2 ; Y = (r 2 + a2 2 0 r )=2;
a = const; 0 = re =2 = const:
122 CHAPTER 7. CONSTRUCTION OF SOLUTIONS. . .

3. Spinor components of the Ricci rotation coeÆcients

 = (r ia cos  ) 1 ;  = cot =(2)3=2 ;
 = ia sin 2 =(2)1=2 ;  =  ;  = Y 2 ;
=  + (r + 0 )=2;  = ia sin =(2)1=2 :
4. Spinor components of the Riemann tensor
2 = = 0 3 :
The metric of the Riemannian solution (7.202) in the coordinates (7.182)
has the form
 
2 0 r 4 0 ra
ds2 = 1 r 2 + a2 cos2  c2 dt2 + r 2 + a2 cos2  sin2 d'cdt
r 2 + a2 cos2  dr 2 (r 2 + a2 cos2  )d 2

r 2 2 0 r + a2 
2 0 ra2
r + a + 2 2 2 sin  sin2 d'2 : (7.203)
2 2 2
r + a cos 

7.7.7 Purely torsional solution

(7.204)
1. Coordinates x = u; x = r; x = ; x = '.
0 1 2 3

2. Components of the Newman-Penrose symbols


0i 0_ = (0; 1; 0; 0); 1i 1_ = (
; Y; 0; a); 0i 1_ = p (ia sin ; 0; 1; icosec );
2


1i 0_ = 0i 1_ ; i00_ = (1; 0; 0; a sin2  ); i11_ =  Y; () 1 ; 0; a sin2 Y ;


i01_ = p ia sin ; 0; ( ) 1 ; i
sin  ; i10_ = i01_ ;
2

= r 2 + a2; Y = (r 2 + a2)=2; a = const:
3. Spinor components of the Ricci rotation coeÆcients
 = (r ia cos  ) 1 ;  = cot =(2)3=2 ;
 = ia sin 2 =(2)1=2 ;  =  ;  = Y 2 ;
=  + r=2;  = ia sin =(2)1=2:
The metric of the Riemannian solution (7.204) in the coordinates (7.182)
has the form
r 2 + a2 cos2 
ds2 = c2 dt2 r 2 + a2 dr (r + a cos  )d
2 2 2 2

(r 2 + a2 ) sin2 d'2 : (7.205)

7.7. SOME PHYSICALLY MEANINGFUL. . . 123

7.7.8 Solution with a variable Coulomb-Newton

interaction and three-dimensional rotation of a
source

(7.206)
1. Coordinates x = u; x = r; x = ; x = '.
0 1 2 3

2. Components of the Newman-Penrose symbols

0i 0_ = (0; 1; 0; 0); 1i 1_ =  (
; Y; 0; a);

0i 1_ = p (ia sin ; 0; 1; icosec ); 1i 0_ = 0i 1_ ;
2
i00_ = (1; 0; 0; a sin2  );


i11_ =  Y; () 1 ; 0; a sin2 Y ;
i01_ = p ia sin ; 0; () 1 ; i
sin  ; i10_ = i01_ ;


2

= r 2 + a2 ; Y = (r 2 + a2 2 0 r )=2;
a = const; 0 = 0 (u):
3. Spinor components of the Ricci rotation coeÆcients
 = (r ia cos  ) 1 ;  = cot =(2)3=2;
 = ia sin 2 =(2)1=2;  =  ;  = Y 2 ;
=  + (r + 0 (u))=2;  = ia sin =(2)1=2;
i _ 0 (u)ra sin 2 
= 1=2
:
2
4. Spinor components of the Riemannian tensor
2 = = 0 (u)3 ;
2 = i _ 0 (u)a sin 2 =(2)3=2 2i _ 0 (u)ra sin 3 =(2)1=2;
4 =  0 (u)ra2 sin2 3 =2 + _ 0 (u)ra2 sin2 4 ;
12 = i _ 0 (u)a sin 2 =(2)3=2;
22 =  (u)ra2 sin2 2 2 =2 _ 0 (u)r 2 2 2 ;
0

The metric of the Riemannian solution (7.206) in the coordinates (7.182)

has the form
 
2 0 (t)r 4 0 (t)ra
ds = 1 r 2 + a2 cos2  c2 dt2 + r 2 + a2 cos2  sin2 d'cdt
2

r 2 + a2 cos2 
dr 2 (r 2 + a2 cos2  )d 2
r 2 2 0 (t)r + a2
 
2 0 (t)ra2
r + a + 2 2 2 sin  sin2 d'2 : (7.207)
2 2 2
r + a cos 
124 CHAPTER 7. CONSTRUCTION OF SOLUTIONS. . .

7.7.9 Solution with electronuclear interaction and

three-dimensional rotation of a source

(7.208)
1. Coordinates x = u; x = r; x = ; x = '.
0 1 2 3

2. Components of the Newman-Penrose symbols

0i 0_ = (0; 1; 0; 0); 1i 1_ =  (; ; 0; a);

0i 1_ = p (ia sin  + 2irN cot ; 0; 1; icosec ); 1i 0_ = 0i 1_ ;

2
 = r 2 + rN2 + a2;  = (r 2 rN2 + a2 re r )=2;
rN = const; a = const; re = const:
3. Spinor components of the Ricci rotation coeÆcients
 = (r + irN ia cos  ) 1 ;  =  0 ;  = 2  0 ;
 = 0 + 2  0 ;  =  0 ;
 = =2 +  0 =2 +  0 =2 + 2  0 ;
= 2 0 + ( 0 0 +  0  0 ) + 2  0  0 ;
0 = re =2 + irN ;
1 1
0 =  0 ;  0 = 4 (2)1=2 cot ;  0 = 2 ia(2)1=2 sin :
4. Spinor components of the Riemannian tensor
2 = = 0 3 :
The nonzero components of the metric tensor gij have the form
guu =  (r 2r rN2 + a2 cos2  );
gur = 1;
gu' = 2rN cos   + 2a sin2  (rer=2 + rN2 ); (7.209)
gr' = a sin2  2rN cos ;
g = r 2 (rN a cos  )2 ;
g'' = (a sin  + 2rN cos  )2  sin2  2 :
2
REFERENCES 125
126 PART 2
Bibliography
[1] Weitzenbock R. Invariantentheorie. Groningen: Noordho, 1923, 320 S.
[2] Weitzenbock R. // Sitzungsber. preuss. Akad. Wiss. Phys.-math. Kl. 1924,
ss. 466-501.
[3] Vitali G. // Atti Soc. ligust. sci. Lett. 1924. Vol. 11, pp. 248-254.
[4] Vitali G. // Ibid. 1925. Vol. 14, pp. 287-291.
[5] Weitzenbock R. // Proc. Knkl. nederl. akad. 1926. Vol. 28, pp. 400-411.
[6] Levi-Civita R. et L. // Math. Ann. 1900. Vol. 54, pp. 125-130.
[7] Schouten J. Ricci-Calculus. B.; Heidelberg: Springer, 1954.
[8] Sintsov D.M. Works on anholonomic geometry. Kiev: Vyshcha shk., 1972
(in Russian).
[9] Cartan E., Schouten J. // Proc. Knkl. nederl. akad. 1926. Vol. 29, pp.
803-810.
[10] Cartan E., Schouten J. // Ibid., pp. 933-938.
[11] Bortolotti E. // Atti Veneto. 1927. Vol. 2, pp. 455-462.
[12] Bortolotti E. // Proc. Knkl. ned. akad. 1927. Vol. 30, pp. 216-311.
[13] Bortolotti E. // Mem. acad. Bologna. 1927. Vol. 30, pp. 45-54.
[14] Bortolotti E. // Rend. Lincei. 1929. Vol. 9, p. 530.
[15] Griss G. Dieratianarianten von Systemen von Vektoren: Diss. Gronin-
gen, 1925. 61 S.
[16] Schouten J. Vorlensugen uber die Theorie der Halbeinfachen Gruppen.
Leiden, 1927. 285 S.
[17] Schouten J. // Math. Ann. 1929. Vol. 102, pp. 244-248.

127
128 PART 2

[18] Eisenhart L. Riemannian geometry. Princeton (N.J.): Univ. press, 1960.

[19] Cartan E. // Math. Ann. 1930. Vol. 102, pp. 698-701.
[20] Thomas T. // Composito math. 1937. Vol. 3, pp. 453-463.
[21] Mayer W., Thomas T. // Ibid. Vol. 5, pp.198-205.
[22] Haimovici A. // Ann. Sci. Univ. Jasay. 1943. Vol. 29, pp. 90-96.
[23] D'Atri J., Nickerson N. // J. Dierential Geometry. 1968. Vol. 2, pp.
393-451.
[24] Wolf J. // Ibid. 1972. Vol.6, p. 317.
[25] Wolf J. // Ibid. Vol.7, pp. 19-26.
[26] Einstein A. // Sitzunsber. preuss. Akad. Wiss. Phys.-math. Kl. 1928, s.
217.
[27] Einstein A. // Ibid., s. 224.
[28] Shipov G.Mathematical foundations of the gauge model of physical vac-
uum, Moscow, 1987. dep. at VINITI N 5326-B87 (in Russian).
[29] Rashevsky P.K. Riemannian geometry and tensor calculus, Moscow,

Nauka, 1964 (in Russian).

[30] Shouten J. Tensor calculus for physicists, Moscow, Nauka, 1965, p. 81 (in
Russian).
[31] Favar J. A course in local dierential geomrtry , Moscow, Foreign Liter-
ature, 1960 (in Russian).
[32] Hayashi K. // Phys. Lett. B. 1977. Vol. 69, N 4, pp. 441-443.
[33] Shipov G. Geometry of absolute parallelism, Part 1, Moscow, 1992,

preprint CISE VENT; N 14, 62 p.

[34] Landau L.D. Lifshits E.M. Mechanics, Moscow , Nauka, 1958 (in Russian).
[35] Hsin-Yang Yeh. // J. Math. Phys. 1974. Vol. 15, N 7, pp. 1085-1095.
[36] Shipov G, Skalsky V.// Proc. of the Intern. Conf. on Dierential Geometry
and Its Applications. Brno, 1989.
[37] Shipov G. Program of Universal relativity and vacuum theory, Moscow,
1988, dep. at VINITI, N 6947-B88 (in Russian).
[38] Penrose R., Rindler V. Structure of space-time, Moscow, Mir, 1972 (in
Russian).
[39] Infeld L., B. der Werden // Akad. Wiss. Phys.-math. Kl. 1933, ss. 380-395.
REFERENCES 129

[40] Newmen E., Penrose R. // J. Math. Phys. 1962. Vol. 3, N 3, pp. 566-587.
[41] Pirani F. Lectures on General Relativity. Vol. 1, 1964.

[42] Witten L. // Phys. Rev. 1959. Vol. 113, pp. 357-362.

[43] Corson E. An introduction to tensors, spinors and relativistic wave equa-
tions. L.: Blakie, 1953.
[44] Carmeli M. // J. Math. Phys. 1970. Vol. 2, pp. 27-28.
[45] Carmeli M. // Lett. nuovo cim. 1970. Vol. 4, pp. 40-46.
[46] Carmeli M. // Phys. Rev. D. 1972. Vol. 5, pp. 5-8.

[47] Shipov G. Geometry of absolute parallelism, Part 2, Moscow, 1992,

preprint CISE VENT N 15, p. 65 (in Russian).

[48] Frolov V. // Trudy FIAN, 1977. Vol. 96, pp. 72-180 (in Russian).
[49] Herrera L. // Lett. nuovo cim. 1978. Vol. 21, pp. 11-14.

[50] Ozsvath J. // J. Math. Phys. 1964. Vol. 6, N 4, pp. 590-611.

[51] Carmeli M., Malin S. // Ann. Phys. 1977. Vol. 103, pp. 208-232.

[52] Carmeli M. // Phys. Rev. D. 1976. Vol. 14, N 6, pp. 1727-1728.

[53] Geroch R., Held A., Penrose R. // J. Math. Phys. 1973. Vol. 14, p. 874.
[54] Penrose R., Rindler V. Spinors and space-time. Vol. 1, Moscow, Mir, 1987
(in Russian).
[55] ShipovG. Geometry of absolute parallelism, Part 3, Moscow, 1992,

preprint CISE VENT N 16, p. 76 (in Russian).

[56] Robinson J., Trautman A. // Phys. Rev. Lett. 1960. Vol. 4, p. 431.
[57] Petrov A.Z. // Uchen. zap. Kazan. univ., 1954. Vol. 114, p. 55 (in Rus-
sian).
[58] Petrov A.Z. New methods in general relativity, Moscow, Nauka, 1966, (in
Russian).
[59] Goldber J., Sachs R. // Acta phys. pol. Suppl. 1962. Vol. 22, pp. 13-18.
[60] Newman E., Unti T. // J. Math. Phys. 1962. Vol. 3, N 3, pp. 891-901.
[61] Newman E., Penrose R. // Ibid. 1963. Vol. 4, N 7, pp. 998-1005.

[62] Newman E., Tamburino L., Unti T. // Ibid., pp. 915-923.

[63] Maia D. // Ibid. 1973. Vol. 14. N 7, pp. 882-887.
[64] Ozsvath J. // Ibid. 1964. Vol. 6, N 4, p.590-611.

[65] Frolov V.P. // TMF, 1974. Vol. 21, N 2. pp. 213-223.

130 A THEORY OF PHYSICAL VACUUM
Contents
5 Geometry of absolute parallelism in vector basis 13
5.1 Object of anholonomicity. Connection of absolute parallelism . . 13
5.2 Covariant dierentiation in A4 geometry. Ricci rotation coeÆcients 15
5.3 Curvature tensor of A4 space . . . . . . . . . . . . . . . . . . . . 19
5.4 Formalismof external forms and the matrix treatment of Cartan's
structural equations of the absolute parallelism geometry . . . . 22
5.5 A4 geometry as a group manifold. Killing-Cartan metric . . . . . 26
5.6 Structural equations of A4 geometry in the form of expanded,
completely geometrized Einstein-Yang-Mills set of equations . . . 30
5.7 Equations of geodesics of A4 spaces . . . . . . . . . . . . . . . . 35
5.8 Structural equations of right and left A4 geometry . . . . . . . . 42
6 The geometry of absolute parallelism in spinor basis 47
6.1 Three main spinor bases of A4 geometry . . . . . . . . . . . . . 47
6.2 Spinor representation of the structural Cartan equations of A4
geometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
6.3 Splitting of structural Cartan equations into irreducible represen-
tations of the group SL(2:C) . . . . . . . . . . . . . . . . . . . . 53
6.4 Carmeli matrices . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
6.5 Component-by-component rendering of structural equations of
A4 geometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
6.6 Connection of structural Cartan equations of A4 geometry with
the NP formalism . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
6.7 Variational principle of derivation of the structural Cartan equa-
tions and the second Bianchi identities of A4 geometry . . . . . . 71
6.8 Decomposition of spinor elds of A4 geometry into irreducible
parts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
6.9 Spinor set of Einstein-Yang-Mills equations . . . . . . . . . . . . 77
6.10 Formalism of two-component spinors . . . . . . . . . . . . . . . . 79
7 Construction of solutions to structural Cartan equations of the
geometry of absolute parallelism 85
7.1 Selection of a frame of reference and specialization of Newman-
Penrose symbols . . . . . . . . . . . . . . . . . . . . . . . . . . . 85

131
132 A THEORY OF PHYSICAL VACUUM

7.2 Specialization of the spinor components of the Ricci rotation co-

eÆcients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
7.3 Specialization of the spinor components of the Riemann tensor . 94
7.4 Construction of the asymptotic behavior of insular-type geometries 96
7.4.1 Radial equations containing derivatives with respect to r 96
7.4.2 Nonradial equations . . . . . . . . . . . . . . . . . . . . . 97
7.4.3 U-derivative equations . . . . . . . . . . . . . . . . . . . . 97
7.5 Classication of solutions to the structural Cartan equations of
the A4 geometry by isometry groups . . . . . . . . . . . . . . . . 106
7.6 A4 geometry with a Schwarzschild-type metric . . . . . . . . . . 108
7.7 Some physically meaningful solutions of the structural Cartan
equations of A4 geometry . . . . . . . . . . . . . . . . . . . . . . 117
7.7.1 Solution with a variable source function . . . . . . . . . . 117
7.7.2 Solution with quark interaction . . . . . . . . . . . . . . . 118
7.7.3 Solution with a short-range (nuclear) interaction . . . . . 119
7.7.4 Solution with an electronuclear interaction . . . . . . . . 119
7.7.5 Solution with electronuclearquark interaction . . . . . . . 120
7.7.6 Solution with Coulomb-Newton interaction and three-dimensional
rotation of a source . . . . . . . . . . . . . . . . . . . . . . 121
7.7.7 Purely torsional solution . . . . . . . . . . . . . . . . . . . 122
7.7.8 Solution with a variable Coulomb-Newton interaction and
three-dimensional rotation of a source . . . . . . . . . . . 123
7.7.9 Solution with electronuclear interaction and three-dimensional
rotation of a source . . . . . . . . . . . . . . . . . . . . . . 124