QUATERNIONS IN APPLIED SCIENCES

A HISTORICAL PERSPECTIVE OF A MATHEMATICAL CONCEPT

HELMUTH R. MALONEK
Departamento de Matemática
Universidade de Aveiro
3810-193 Aveiro
Portugal
e-mail: [email protected]

After more than hundred years of arguments in favor and against quaternions,
of exciting odysseys with new insights as well as disillusions about their useful-
ness the mathematical world saw in the last 40 years a burst in the application of
quaternions and its generalizations in almost all disciplines that are dealing with
problems in more than two dimensions. Our aim is to sketch some ideas - neces-
sarily in a very concise and far from being exhaustive manner - which contributed
to the picture of the recent development. With the help of some historical reminis-
cences we firstly try to draw attention to quaternions as a special case of Clifford
Algebras which play the role of a unifying language in the Babylon of several dif-
ferent mathematical languages. Secondly, we refer to the use of quaternions as
a tool for modelling problems and at the same time for simplifying the algebraic
calculus in almost all applied sciences. Finally, we intend to show that quaternions
in combination with classical and modern analytic methods are a powerful tool for
solving concrete problems thereby giving origin to the development of Quaternionic
Analysis and, more general, of Clifford Analysis.

Could anything be simpler or more satisfactory? Don’t you feel, as well as

think, that we are on a right track, and shall be thanked hereafter. Never
mind when.
W. R. Hamilton 1859
...there will come a time when these ideas, perhaps in a new form, will arise
anew and will enter into living communication with contemporary develop-
ments...
H. G. Grassmann 1861

1 Instead of an Introduction

The NOTICES OF THE AMS, VOL. 48 (2001), NUMBER 4, 411-412 announced

the solution of a problem in fluid mechanics that had been open and actively in-
vestigated for a hundred years.

The Ruth Lyttle Satter Prize in Mathematics is awarded to Sijue Wu for

her work on a long-standing problem in the water wave equation, in par-
ticular for the results in her papers (1) Well- posedness in Sobolev spaces
of the full water wave problem in 2-D, Invent. Math. 130 (1997), 39-72;
and (2) Well-posedness in Sobolev spaces of the full water wave problem
in 3-D, J. Amer. Math. Soc. 12, no. 2 (1999), 445-495. By applying
tools from harmonic analysis (singular integrals and Clifford algebra), she
proves that the Taylor sign condition always holds and that there exists a
unique solution to the water wave equations for a finite time interval when

1
 the initial wave profile is a Jordan surface.

The recognition of this outstanding contribution has been confirmed on the oc-
casion of the Second International Congress of Chinese Mathematicians in Taipei
in December, 2001. Sijue Wu was awarded the 2001 Morningside Silver Medal in
Mathematics for her work on water wave problems.
An abstract from the Technical Report Server of the Johnson Space Center reveals
in 1995 that

...To compensate for the required iteration methodology, all reference

frame change definitions and calculations are performed with quaternions.
Quaternion algebra significantly reduces the computational time required
for the accurate determination of shadow terminator points.
[C. R. Ortiz Longo and St. L. Rickman, Method for the Calculation of
Spacecraft Umbra and Penumbra Shadow Terminator Points]

In September 2002 the UNDERGRADUATE STUDENT MANUAL of the De-

partment of Mechanical Engineering and Applied Mechanics at the University of
Pennsylvania mentioned :

528. Advanced Kinematics. (M) Prerequisites: Multivariate calculus, in-

troductory abstract algebra, mathematical maturity.
Differential geometry, Lie groups and rigid body kinematics; Lie algebra,
screws, quaternions and dual number algebra; geometry of curves and ruled
surfaces; trajectory generation and motion planning; applications will be
to robotics and spatial mechanisms.

In a typical abstract of a contribution on quaternions to a recent conference in

mathematics and its applications (GAMM) the following remarks can be found:

... we consider applications of quaternions (a subcase of the more gen-

eral Clifford Algebras) to different problems of orientation in engineer-
ing. It is shown that many problems from different engineering areas such
as computer vision, robotics, navigation, photogrammetry, etc. have the
same mathematical background and can be formulated as quaternion op-
timization problems. ... Thus, by using quaternions we have en elegant
mathematical method for solutions of many complicated problems in dif-
ferent areas of engineering. The quaternion approach allows to clarify the
essence of problems and to simplify numerical calculations.

An important indicator for the dynamic in the field is the publication of books.
Some examples from the last 6 years are

• Applications of Geometric Algebra in Computer Science and En-

gineering (Dorst, L., Doran, Ch., Lasenby, J. (eds.), Birkhäuser,
2002)[11]
• Geometric Algebra with Applications in Science and Engineering
(Corrochano, E. B. and G. Sobczyk (eds.) Birkhäuser, 2001)[6]

2
 • Geometric Computing with Clifford Algebras Theoretical Founda-
tions and Applications in Computer Vision and Robotics (Sommer,
G. (ed.), Springer, 2001)[41]
• Clifford Algebras and their Applications in Mathematical Physics,
Vol. 1 Algebra and Physics (AbÃlamowicz, R., Fauser, B. (eds.)) Vol.
2 Clifford Analysis, (Ryan, J., Sprößig, W.(eds.) Birkhäuser, 2000)[1]
• Electrodynamics: A Modern Geometric Approach, (Baylis W. E.,
Birkhäuser, 1999)[3]
• Quaternions and Rotations Sequences, (Kuipers, J. B., Princeton
Univ. Press, 1998)[34]
• Quaternionic and Clifford Calculus for Physicists and Engineers
(Gürlebeck, K.,Sprößig, W., John Wiley &. Sons, 1997)[20]

To complete this list we still quote a notice on the history of the Mathematics
Department at Union College (USA):

Gillespie not only stressed the importance of mathematics in engineering,

he began a tradition of emphasizing the humanities for the engineers as
well; this is a tradition that the College has wisely preserved to this day, to
its enormous benefit. The Catalogue in 1882-83 listed, among many other
things, a course in quaternions, which at the time was right at the fron-
tier of mathematical research, and involved some of the biggest names in
American mathematics (for example, Sylvester on one side and J. Willard
Gibbs of Yale on the other).

The aforementioned facts speak for themselves. As 120 years ago quaternions and
their generalizations are present in recent mathematical research, in applications as
well as in teaching. Surely, they are not an esoteric topic for pure mathematicians,
but rather a mathematical tool for engineering. Nevertheless, is it not strange that
a theory which was created for applications 160 years ago could stay (with only
very few exceptions) almost dormant for about 120 years?

2 Clifford Analysis and Geometric Algebras - Quaternions on the

Right Track

Hamilton believed that he was ”on the right track”. But it is a historical fact that
in the very beginning the true nature of quaternions was misunderstood, also by
Hamilton, and the consequences were ”Missed opportunities” (F. Dyson,[12]) which
caused a delay of about 40 years in what concerns their application in Physics. In
pure mathematics the progress was never completely stopped, but mainly going on
in Number Theory and Abstract Algebra with controversial academic discussions
about the right understanding of quaternions. In engineering the vector, tensor
and matrix calculus dominate everywhere, at least since the beginning of the 20th
century.
From the historical point of view it is very difficult to do justice to the
pioneers of that time, but it seems to us that two events marked significantly some

3
changes. One was the newly started research on hypercomplex function theory by
Iftimie [32](in the tradition of works done by his romanian co-patriots Moisil and
Toedorescu in 1931 [38]) and by Delanghe [8](following the line of Fueter’s school
in Zurich between 1930 and 1950). But whereas the motivation of Fueter [13]came
from number theoretic problems, Delanghe and his soon growing Gent school on
Clifford Analysis concentrated on problems arising from harmonic analysis, coming
much closer to significant applications in Physics. The other event was initiated by
the physicist D. Hestenes who had worked for NASA and published ”Space-time
algebra”[26] and ”Multivector calculus” [27] in 1966 resp. 1968. His work was
mainly concerned with physical problems expressed in the language of Geometry.
Following Clifford’s original idea he coined the term ”Geometric Algebra” for his
approach to the use of Clifford algebras in applications. Clifford Analysis and
Geometric Algebras are nowadays synonyms for the two main ways of dealing with
Clifford algebras: the first with emphasis in the analytic theory as generalization
of Complex Analysis in a wide sense, including modelling and numerical methods,
the second one with emphasis in geometric-algebraic models of very diversified
physical or technical problems.

Finally, it was less than 20 years ago, in 1985, that the first international confer-
ence on ”Clifford Algebras and Their Applications in Mathematical Physics” took
place in Kent, Canterbury, U.K., joining mathematicians, physicist and engineers.
Clifford Analysis and Geometric Algebras met each other on that occasion. D.
Hestenes [28]noticed in his address to the conference:

This first international workshop on Clifford Algebras testifies to an in-

creasing awareness of the importance of Clifford Algebras in mathematics
and physics. Nevertheless, Clifford Algebra is still regarded as a narrow
mathematical speciality , and few mathematicians or physicists are likely
to characterize Clifford Algebra as merely the algebra of a quadratic form,
while the physicists are likely to regard it as a specialized matrix algebra.
...
The fact that Clifford Algebra keeps popping up in different places through-
out mathematics and physics shows that it has a universal significance
transcending narrow specialities.

Meanwhile a lot of disciplines in science and engineering got increasingly inter-

ested and contribute intensively to the spectra of applications of quaternions and
their generalizations. We will mention several of them but, of course, are not able
to draw a complete picture. The books we mentioned in the preceding section can
serve as a guide for getting some overview about recent developments.
To get closer to the general perspective and the role of quaternions as a math-
ematical concept in theory and applications we will review some basic algebraic
relations and few analytic methods in more detail. We begin to recall some ques-
tions that arose in the very beginning of the creation of quaternions which needed
a long time to be understood and re-interpreted.

4
3 Hamilton - his Discovery and his Obsession

Though the early history of quaternions is more or less well known, several basic
facts are best explained in their historical context, particularly their origin as a
tool, not as a theoretical concept. The discovery of quaternions by the Royal As-
tronomer of Ireland, William Rowan Hamilton [22], on the 16th of October 1843
was motivated by the hope to create a type of hypercomplex numbers related to
the three dimensional space of our visual intuition like complex numbers are re-
lated to the plane. At that time, one of the most striking facts about the complex
numbers was their geometric visualization (Wessel, Argand, Gauss) and the dis-
covery that the simple algebraic operation of multiplication could be interpreted
in terms of rotations in the plan. It became an intensely studied question whether
one could discover other number systems which would model rotations in three
dimensional space. Hamilton succeeded to go an important (now seemingly triv-
ial) step forward to the solution. His formal definition of complex numbers as
ordered pairs of real numbers (1835) suggested to him the idea to attack the prob-
lem as an algebraic one for ordered triplets (α, β, γ) of real numbers combined to
z = α + βi + γj with i2 = j 2 = −1. But it took him almost ten years to under-
stand that without a fourth dimension and without dropping the commutativity of
multiplication no such system exists. Indeed, in a letter to J. Graves [23](17th of
October 1843) he wrote:

...I made therefore ij = k, ji = −k, reserving to myself to inquire whether

k was 0 or not...

and also

...and there dawned on me the notion that we must admit, in some sense,
a fourth dimension of space for the purpose of calculating with triplets...

After having arrived to the basis representation of his quaternions as expressions

of the form
q = q0 + q1 i + q2 j + q3 k, qm ∈ IR, m = 0, 3 with i2 = j2 = k2 = ijk = −1, (1)
Hamilton showed very carefully that besides commutativity all other properties
that characterize quaternions as a number system (in particular the associativity
of multiplication) are fulfilled. Since the inverse of a nonzero quaternion q is given
by
q −1 = (q02 + q12 + q22 + q32 )−1 (q0 − q1 i − q2 j − q3 k) = |q|−2 q̄ (2)
Hamilton [25] constructed with the set IH of all quaternions a real finite-dimensional
associative division algebra unknown so far, which later was recognized by Frobenius
(1877) as the only one besides IR and C. In fact, his discovery heralded the golden
age of algebra when instead of algebraic equations the study of algebraic structures
became the main concern. But it seems that Hamilton’s concern was not this
essential contribution to modern algebra.
In fact, Hamilton was already at that time a recognized physicist working in
geometric optics on extremal principles and successfully extending these ideas to

5
dynamics in 1834/35. With the introduction of the principle of least action, the
Hamilton function and his canonical equations of motion, he had already reached
the Hall of Fame, having been knighted in 1835 and elected for President of the
Royal Irish Academy in 1837. This explains why his prevailing attitude towards
quaternions was their applicability, their geometric (physical) nature, not their
algebraic properties. From the very beginning of his trials he was obsessed by the
quixotic idea that quaternions would play a key role in physics, being on a par with
the creation of the infinitesimal calculus. Consequently, for the last twenty years
of his life, Hamilton concentrated all his power on the study of quaternions.

Note: It should be mentioned that, at first sight, the quaternion formalism

might seem awkward to a physicist or engineer, since the square of the unit vectors
i, j and k are negative; besides, invoking a fourth dimension beyond our ability
of visualization for treating 3-dimensional problems also looks rather strange. In-
deed, many of the obstacles in gaining a firm reputation for easy applicability of
quaternions were (and still are) caused by these ”strange” properties. Hamilton’s
own struggle seems to show that he was aware of it. But the man who succeeded
to explain the secrets of the new number system with relations to dimension three
and four was a man of a new generation, born after 1843. His name was William
Kingdom Clifford (1845-1879).

4 Quaternions and Vector Calculus

The most natural example of an elementary tool that is widely used in geome-
try, physics and any technical science, without showing directly its relationship to
quaternions, is vector calculus. Nevertheless, and almost forgotten, vector alge-
bra as well as vector analysis have their origin in the theory of quaternions, i.e.
historically quaternions were the first.
It is very curious, may be in some sense ironic and even of philosophical signif-
icance, that the appearance of quaternions also marked the beginning of modern
vector analysis, which later on prevailed over the use of quaternions. The fact that
Hamilton started with triplets and ended up with quaternions implied immediately
his concern about the special role of q = q1 i + q2 j + q3 k ∈ ImIH which he called
vector in the sense of the similar term radius vector that had been used for many
years before. Hamilton wrote:
Regarded from a geometrical point of view, this algebraically imaginary
part of a quaternion has thus so natural and simple a signification or
representation in space, that the difficulty is transferred to the algebraically
real part; and we tempted to ask what this last can denote in geometry, or
what in space might have suggested it.
For the algebraically real part he introduced the word scalar part or simply the
scalar of q and prefixed both components of q by V resp. S, i.e. q = Sq + V q.
The use of S and V applied to the product of two quaternions being ”vectors”
α = xi + yj + zk and α0 = x0 i + y 0 j + z 0 k then leads to
Sαα0 = −(xx0 +yy 0 +zz 0 ) resp. V aa0 = i(yz 0 −zy 0 )+j(zx0 −xz 0 )+k(xy−0 yx0 ). (3)

6
But formula (3) shows that the quaternion product is the sum of the modern vector
(cross) product and the negative of the modern scalar (dot) product of α and α0 .
The present formulation of vector algebra comes from this quaternion product of
triplets/vectors in the sense of Hamilton, extracted by Gibbs 1881-84 and first
published in 1901.
With the introduction of the Nabla-operator ∇ in the form
∂ ∂ ∂
∇=i +j +k (4)
∂x ∂y ∂z
in 1846/1847, Hamilton also invented the other essential technical ingredient for
the vector calculus: the vector differential operator which is used to describe the
gradient of a scalar function as well as the divergence and the curl of a vector valued
function.
Long time after Hamilton the vector calculus found in Gibbs and Heaviside their
most prominent promoters. Remarkable and surely not fair, that both strongly
tried to deny that inheritance as it is described in Crowe’s book on the History
of Vector Analysis [7]. Perhaps because of the objective of this book (expressed
by its title) the role of another, probably the most important, mathematician in
this story, W. K. Clifford, is only described in there as a ”transition figure”. In
fact, Clifford’s insights in the relationship between Hamilton’s quaternions and
Grassmann’s ”algebra of extensions” are the key to a fair appreciation of Hamilton’s
work as well as the milestone for the progress in this field of algebra.
But before we discuss the role of Clifford we have to mention the use of quater-
nions for describing rotations (one of the most important procedures in all tech-
nical applications). Hamilton soon found out how the desired description of rota-
tions in IR3 can be related to a vector (nowadays also called ”pure quaternion”)
a = xi + yj + zk. He first wrote a rotation a 7→ a1 in the form a1 = ua where
u ∈ IH, |u| = 1, is a unit quaternion. But in such a rotation, the imaginary part u
of u (which describes the axis of the rotation) had to be perpendicular to the vector
a. Altmann extensively discussed this insufficiency in Hamilton’s early approach to
rotations in ([2]). Of course, this was the form that Hamilton expected from the
complex multiplication and it is also the form of the matrix representation of rota-
tions in IR3 . Indeed, using the 3 × 3-rotation matrix U corresponding to a rotation
of ϕ about the axis given by uT there holds a1 = U a.
But the concept of a matrix was also motivated by quaternions: matrix
algebra was introduced by Cayley only in 1858. Cayley, seriously concerned
with Hamilton’s creation of quaternions, published in 1845 the right formula for
rotations represented by quaternions (he assigns the priority to Hamilton):

Every rotation (properly orthogonal mapping) of a ∈ IR3 has the form

ϕ u ϕ
a 7→ uau−1 , where u = cos + sin (5)
2 ϕ 2
runs through all nonzero quaternions and ϕ = |u|.
Hamilton’s particular problem was to explain the appearance of the half of the
rotation angle in this general rotation formula. According to (2) this means that a

7
rotation of a ∈ IR3 is realized by multiplication with a unit quaternion from the left
and its inverse from the right (almost like rotations in the complex plane, except
that there commutativity allows the multiplication with exp(iϕ) from one side
only). Notice that two unit quaternions u and −u represent the same rotation and
that the composition of two rotations in IR3 in general is not commutative (this,
of course, has its true counterpart in the non-commutativity of the quaternionic
multiplication). In modern group theoretic terms (cf. the next section) this means
that unit quaternions are elements of the double covering group Spin(3) of the
rotation group SO(3) in IR3 . Finally we should notice that later on, in 1855 Cayley
remarked that every properly orthogonal mapping of IR4 ∼ = IH has the form

q 7→ aqb

where a, b independently of each other run through all unit quaternions.

5 Clifford’s Geometric Algebra

Now we will see how a deeper theoretical background on the algebra of quaternions
allows to benefit more from their properties. Thanks to Clifford [5] this approach
started soon after Hamilton, and seems to be still in progress. The variety of
Clifford algebra representations is very big and a lot of different approaches exist.
Specially in theoretical physics, the use of Clifford algebras is meanwhile abundant
and the number of different possibilities of writing Maxwell’s equations by using
Clifford algebras is very high. The same concerns the equations of motion of the
spinning electron in quantum mechanics, first obtained by Dirac when trying to
linearize the Klein-Gordon equation [10].
Clifford has been considered as one of the four mathematicians that very soon
understood the ideas of Grassmann, after their publication in the confusing work
”Die lineale Ausdehnungslehre - ein neuer Zweig der Mathematik” [18]. He suc-
ceeded to combine in his ”Geometric Algebra” the V ideas of both, Hamilton and
Grassmann. Indeed, in 1844 the exterior algebra IR3 of the linear space IR3 was
constructed by Grassmann with the basis

1
e1 , e2 , e3 ,
e1 ∧ e2 , e 1 ∧ e3 , e2 ∧ e3
e1 ∧ e2 ∧ e3

satisfying the multiplication rules

(i) ei ∧ ej = −ej ∧ ei for i 6= j, (6)

(ii) ei ∧ ei = 0.

The Grassmann exterior algebra has no inner product and does not require
a metric. Therefore Clifford, motivated by Hamilton’s ideas, introduced in 1878
a new product where metric relations for products of vectors (realized by areas,
volumes etc.) also have place. He kept the first multiplication rule and replaced

8
the second by

ei ej = −ei ej , (7)
ei ei = −1.

In the concrete case of IR2 this led him to the basis

1 the scalar
e1 , e2 , vectors
e1 e2 , bivector or area element

of the real associative 4-dimensional Clifford algebra C`0,2 . The dimension grading
introduces a multi-vector structure and the multiplication rules show that C`0,2 ∼ =
IH, i.e. Clifford obtained an algebra which is isomorphic to Hamilton’s quaternions
with the basis {1, e1 = i, e2 = j, e1 e2 = k}. But from the geometric point of
view this approach was not satisfactory, because the nature of the basis elements
{1, e1 = i, e2 = j, e1 e2 = k} seems different (2 vectors and 1 bivector) and the
square of an element of the underlying vector space IR2 remains negative. Clifford
found only later the way out to an algebra which fulfills all demands. His work was
published posthumously in 1882. Therefore he changed the multiplication rules to

ei ej = −ei ej , (8)
ei ei = +1.

In the concrete case of IR3 this led him to the basis

1 the scalar
e1 , e 2 , e3 , vectors
e1 e2 , e1 e3 , e2 e3 bivectors
e1 e2 e3 trivector or volume element.

Here the quaternions can be recognized as isomorphic to the even subalgebra C`+ 3,0
constituted by scalars and bivectors (visualized by oriented areas) of the real as-
sociative 8-dimensional Clifford algebra C`3,0 according to the following correspon-
dences:

i → −e2 e3 , j → −e3 e1 , k → −e1 e2

Indeed, for example i2 = (−e2 e3 )2 = −e2 e3 e3 e2 = −e22 e23 = −1 and ij =

(−e2 e3 )(−e3 e1 ) = −e2 e1 = k as well as ijk = −e2 e3 e3 e1 e1 e2 = −1.
In this case all basis elements i, j, k have an interpretation as basic bivectors.
Therefore the set of pure quaternions ImIH as the set of real linear combinations
of i, j, k geometrically represents the set of bivectors and not the set of vectors in
the sense of radius vectors like Hamilton had supposed. This explains also, why
Hamilton met problems when trying to explain the appearance of the half of the
rotation angle in the general formula of rotation (see [2]).
The embedding of IH in the even subalgebra C`+ 3,0 of the Clifford algebra C`3,0 is
also of relevance to the relation between the quaternionic product of pure quater-
nions and the dot- and cross-products of vector analysis. Indeed, it follows from

9
(3) that
(a1 i + a2 j + a3 k)(b1 i + b2 j + b3 k) = −(a1 b1 + a2 b2 + a3 b3 ) + (a2 b3 − a3 b2 )i
+(a3 b1 − a1 b3 )j + +(a1 b2 − a2 b2 )k (9)
= −(a · b) + (a × b)
but the Clifford product between two arbitrary vectors a, b ∈ C`3,0 ∼
= IH is obtained
as
ab = (a1 e1 + a2 e2 + a3 e3 )(b1 e1 + b2 e2 + b3 e3 ) =
(a1 b1 + a2 b2 + a3 b3 ) + (a2 b3 − a3 b2 )e2 e3 + (a3 b1 − a1 b3 )e3 e1 + (a1 b2 − a2 b2 )e1 e2
= (a · b) + (a ∧ b)
= (a · b) + (a × b)e1 e2 e3
The last formula shows that

1. The Clifford product of two vectors belongs to C`3,0 and is a sum of the usual
dot-product (scalar product) of vector analysis, as its scalar part, and the
corresponding cross-product (vector-product) times the so called pseudo-scalar
e1 e2 e3 , as the bivector part.

2. The Clifford product of two vectors from C`3,0 is the sum of an inner (scalar-
valued) and an outer product (bivector-valued). Since the inner product com-
mutes and the outer product anti-commutes both can be derived from the
Clifford product in the form
1 1
(a · b) = (ab + ba)resp.(a ∧ b) = (ab − ba)
2 2
Grassmann and Clifford belong to first group of mathematicians who overcame
the barriers of the 3-dimensional space of our intuition. Clifford introduces his new
algebras not only in IR3 but in IRn . In the modern treatment of Clifford algebras it
is usual to permit that some of the elements of the orthonormal basis e1 , e2 , . . . en
of IRn = IRp+q have positive squares whereas other have negative squares, corre-
sponding to the consideration of Euclidean or Pseudo-Euclidean spaces (like the
Minkowski space IR1+3 ). The multiplication rules
ei ej = −ei ej , where ei ej = ei ∧ ej (10)
ei ei = 1, i = 1, 2, . . . p (11)
ej ej = −1, j = 1, 2, . . . q, p + q = n,
again produce a real 2n -dimensional basis of the corresponding Clifford algebra
C`p,q given by
1
e1 , e2 , · · · en ,
e1 e2 , e 1 e3 , · · · en−1 en ,
e1 e2 e3 , ··· ···
··· ··· ···
e1 e2 e3 · · · en .

10
 Note: It has been shown that Clifford himself studied already two different
types of algebras which allow a geometric interpretation that goes behind the usual
vector concept and makes Grassmann’s extension ideas (in form of the outer product
and the concept of bivectors) manifest. In the beginning of this section we also
mentioned that in applications several variants of Clifford algebras are in use. The
most accepted for practical purpose in the 3-dimensional Euclidean space E3 seems
to be a variant that was proposed almost 40 years ago by D. Hestenes [28]. He
called it simply the ”Geometric Algebra” and based it on G3 := C`3,0 with the
distinguished element i = e1 e2 e3 = e1 ∧ e2 ∧ e3 which is called pseudoscalar.
Given a vector x in G3 , he defined
E3 := {x : x ∧ i = 0}.
The even subalgebra of scalars and bivectors G3+ is called spinor algebra and through
the isomorphism given by the correspondences
e2 e3 → i, e3 e1 → j, e1 e2 → k
(notice that in this case ijk ∼
= e2 e3 e3 e1 e1 e2 = 1) the Quaternion algebra is inter-
preted as a Spinor algebra. This is motivated by the spinor representation theory
of the rotation groups SO(n) which first systematically has been developed in 1913
by Elie Cartan.
It is worth noticing that Freeman J. Dyson [12], who delivered in 1972 the
famous annual J.W.Gibbs Lectures (Bull. AMS, Vol. 78 (1972), 635-652) discussed
, among other themes, under the title ”Missed Opportunities”, also the lost time
when quaternionists and anti-quaternionists were fighting against each other. He
said:
Gibbs had not really succeded in unifying the notions of quaternion and
vector. On the contrary, by putting the two notions side by side he had
made explicit the lack of any real compatibility between them. His lecture
On multiple algebra[17] ought to have suggested to any attentive mathe-
matician the question,
”How can it happen that the properties of three-dimensional space are
represented equally well by two quite different and incompatible algebraic
structures?”
If this question had once been clearly asked, the answer would almost cer-
tainly have been forthcoming. And the answer would have led inevitably to
a complete theory of the single valued and double-valued representations
of the three-dimensional rotation group.The vectors are the simplest non-
trivial single-valued representation, and the quaternions are the simplest
double-valued representation. Also, the quaternions are the prototype of
what later were called spinor representations. The development of spinor
representations, which was actually begun by Elie Cartan in 1913 and com-
pleted during the 1930’s with substantial help from the physicists Pauli and
Dirac, might have been accelerated by approximately 40 years. It is impos-
sible to say what effects such an accelerated development would have had
on other branches of pure mathematics, but the effects could hardly have
failed to be substantial.

11
W. Pauli (1927) and P.A.M. Dirac rediscovered quaternions by their matrix repre-
sentations in M at(2, C) ∼
= C`3,0 resp. M at(4, C) ∼= C ⊗ C`1,3 . Later on the mathe-
maticians R. Brauer and H. Weyl (1935) as well as C. Chevalley (1954) contributed
essentially to this important field of research in quantum theory and specialists in
Physics and Clifford algebras are still very active in this field [1].

6 Trends in Geometric Algebra

Since about 40 years, the field of Geometric Algebra has been increasing and broad-
ening itself in such a way that it is no longer possible to make a simple classification
or analysis of its trends, even of the main trends regarding the concrete objects of
research and application. It seems obvious that the natural and profound relation-
ship to practical computer sciences, in the sense of computational geometry, turned
out to be the main impulse for the development of new applications of Geomet-
ric Algebras. But also the re-interpretation of classical theories, like for instance
the work of Hestenes on ”Point groups and space groups” [29] which implies a
better understanding of molecular modelling and crystallography still plays an im-
portant role. In this sense Hestenes’ remarks at the first international conference
on ”Clifford Algebras and Their Applications in Mathematical Physics” should be
remembered. He noticed that several different systems which provide similar geo-
metric concepts, need to have a common language. Without trying to reduce or to
question the advances and the usefulness of disciplines like
Vector Analysis Tensor Analysis
Matrix Algebra Clifford Algebra
Differential Forms Coordinate Geometry
Synthetic Geometry Grassmann Algebra
Spinor Calculus Multivector Algebra,
all together constitute a set of highly redundant theories which sometimes need
enormous technical skills. It seems to be rather evident that the coordinate-free
tools provided by Geometric Algebra could serve in some sense as a unifying
language for several different mathematical languages that are used in different
fields of application. Further success will be the guarantee for progress in this
direction.

What concerns modern trends we should mention the following. It is well known
that the use of quaternions for instance in the fields of
- aircraft orientation,
- spacecraft stabilization,
and other areas of relevance for defense were a well hidden secret during the cold
war on both sides of the iron curtain.
Meanwhile one can say that mathematical models based on quaternions have
been very much appreciated in all high technologies with need of calculations in
real time.

Software development for video games, for which quaternions allow efficient com-
putations with minimal storage for smooth rotations of solids is a field of enjoyment

12
which seems to increase without any limitations.
Another examples are the advances in
- robotics,
- computer vision,
- virtual reality,
- calculations in crystallography,
- electrical engineering,
- quantum information processing by nuclear magnetic resonance,
- neural computing etc.
where meanwhile the use of quaternions or derived algebraic structures has become
indispensable.

7 New Horizons - Clifford Analysis

Not more than 20 years have passed since the book [4]of Brackx, Delanghe and
Sommen ”Clifford Analysis” from 1982 coined the name of a new discipline. In
Kluwer’s ”Encyclopedia of Mathematics” from 1997 one reads
Clifford analysis studies functions with values in a Clifford algebra, and, as
such, is a direct generalization to higher dimensions of the classical theory
of functions of one complex variable .... It has its roots in quaternionic
analysis, which was developed from the 1920s onwards as an, albeit modest,
counterpart of the theory of two complex variables. The latter was to evolve
into the vast and rich theory of several complex variables. The former
gained renewed interest in the 1950s and led to hypercomplex function
theory (cf. also Hypercomplex functions), renamed Clifford analysis in the
1980s, when it grew into an autonomous discipline.
Indeed, almost at the same time when Pauli and Dirac studied Schrödinger’s
and Klein-Gordon’s equations with the help of quaternion algebras the quaternions
appeared also in the focus of a new type of research. It is not so well known that
the inspiration for this came from a number theoretic problem.

7.1 Fueter’s approach

A great part of number theoretic research on the quaternion algebra as associative
division algebra had already been realized by Hurwitz [31]. The same time, in
1920, an important problem in the main stream of number theory, influenced by
Hilbert’s twelfth problem, had just been solved by the Japanese mathematician
Tagaki. It was related to the class formula for Abelian number fields over an
imaginary quadratic field and the so-called problem of complex multiplication. The
Swiss mathematician Rudolf Fueter (1880-1950), as a former student of Hilbert
[30], was interested in the generalization of this problem to the quaternionic case.
Since the main tools in this field of analytic number theory are results from complex
function theory he tried (since 1928) to develop a hypercomplex function theory by
considering quaternion valued functions f = f (z) of a quaternion variable z ∈ IH
[13, 14, 16]. (In the last decade of his life he also considered the general case of
C`0,n -valued functions (cf. [15]).

13
 Due to the fact that a quaternion z ∈ IH can be represented by a pair of two
complex variables z1 = x0 + x1 i and z2 = x2 + x3 i in the form
z = x0 + x1 i + x2 j + x3 k = z1 + z2 j.
his work was also a contribution to the very intensive discussion in the 1920th
about the ”right” way for generalizing holomorphic function to higher dimensions.
Of course, the algebraic properties of the quaternion algebra lead to another gen-
eralization than the consideration of a complex valued functions of two or more
complex variables. In this sense, the challenge for Fueter to develop quaternionic
analysis (or hypercomplex function theory as he called it) was twofold.
The fact that the quaternion form a division algebra promised that it simply
could be sufficient to follow a generalization of Cauchy’s approach to holomorphic
complex functions by demanding the existence of a quaternionic differential quotient
as the limit of the form
³ ´³ ´−1
lim f (z + ∆z) − f (z) ∆z . (12)
∆z→0
or
³ ´−1 ³ ´
lim ∆z f (z + ∆z) − f (z) . (13)
∆z→0

But in fact, limits of the expressions (12) or (13) independent from the direction
of convergence, exist only for right resp. left linear functions of a quaternionic
variable. This has rigorously been proven by several author’s independently from
Fueter’s research and until many years later, even until the 1990’s! But Fueter
was aware of a paper of Scheffers [40] where such difficulties had been announced
(not well proved), and used another approach for defining generalized holomorphic
functions (which he called regular).
Following Riemann’s approach in the complex case, he defined quaternion valued
regular functions f (z) = f0 (z) + f1 (z)i + f2 (z)j + f3 (z)k as belonging to the kernel
of the quaternionic Cauchy-Riemann operator
∂ ∂ ∂ ∂
D := + i+ j+ k. (14)
∂x0 ∂x1 ∂x2 ∂x3
This means that f is called a regular from the right function if f D = 0 resp. a
regular from the left function if Df = 0.
Nowadays in Clifford Analysis regular functions are called monogenic (some-
times also hypercomplex holomorphic, Clifford holomorphic etc.).
The far reaching analogy between complex function theory and quaternionic
function theory can be illustrated by some of the basic properties of monogenic
functions.
Indeed, using a quaternion valued surface-element of the form
dZ := dx1 dx2 dx3 − dx0 dx2 dx3 i + dx0 dx1 dx3 j − dx0 dx1 dx2 k
and ω = f (z)dZg(z) with f, g ∈ C 1 (Ω) he noticed that it is possible to derive
the quaternionic form of Stokes’ formula over a 4-dimensional positively oriented

14
domain Ω as
Z Z
f (z)dσg(z) = (f Dg + f Dg)dV (15)
∂Ω Ω

where dV stands for the volume element dV := dx0 dx1 dx2 dx3 .
This hypercomplex form of Stokes’ formula in IR4 immediately suggests the

Generalized Cauchy theorem

Let f be a function monogenic from the right (i.e. f D = 0) and g be a function
monogenic from the left (i.e. Dg = 0) in the 4-dimensional positively oriented
domain with boundary Ω than
Z
f (z)dσg(z) = 0. (16)
∂Ω

Other important properties of monogenic functions follow like in the complex

plane from the fact that the Laplace operator can be factorized, namely as DD =
DD = ∆4 , where
∂ ∂ ∂ ∂
D := − i− j− k
∂x0 ∂x1 ∂x2 ∂x3
denotes the conjugated Cauchy-Riemann operator.
This shows that monogenic functions and their components are harmonic func-
tions. Furthermore, the real analyticity of harmonic functions together with special
functions methods (Legendre polynomials or Gegenbauer polynomials) can be used
for defining generalized power series. A generalization of the concept of the areolar
derivative in the sense of Pompeiu,[39] which relies on measure theoretic relations
between higher-dimensional volume and surface-integrals, enables to show that 12 Df
can be considered as the generalized hypercomplex derivative of the monogenic func-
tion f (see [19]). This answered the question about the existence of a generalized
Cauchy-approach as mentioned in the beginning of this section. In fact, it is also
a hint to the fact that hypercomplex analysis can be considered as function theory
in co-dimension 1.
Examples:
Using the shorthand notations e1 := i, e2 := j, e3 := k there holds

1. ([35]) fk (z) = zk := xk − x0 ek = − 12 [zek + ek z], (k = 1, . . . , n) are right- and

left-regular (totally regular variables). Identifying ek ∼= i they are 3 copies of
the complex variable z multiplied by −ek , i.e. zk ∈ Ck := −ek C.

2. The identity f (z) = z is not monogenic since Df = f D = −2. Powers of z, i.e.

f (z) = z n as well as simple products like zj · zk , j 6= k are also not monogenic.

3. [36] The symmetrized products 12 [zj · zk + zk · zj ] = xj xk − x0 xk ej − x0 xj ek

are right- and left-monogenic. The same is true for m-fold symmetric products
which are homogeneous of degree m and can serve as a complete basis in
questions of interpolation or approximation.

15
 4. If ω3 stands for the area of the unit sphere S 3 in IR4 then
1 z−ζ
E(z, ζ) =
ω3 |z − ζ|4
generalizes the Cauchy kernel
1 1
C(z, ζ) =
2πi z − ζ
in the plane, more precisely E(z, ζ) is the right (resp. left) fundamental solution
(in the distributional sense) of the operator D in the unit ball B(1, ζ) = {z :
|z − ζ| ≤ 1}. With other words, applying (15) with f or g chosen equal to
E(z, ζ) a generalized Cauchy integral formula of the form
Z Z
f (z) = f (z)dσE(z, ζ) resp. g(z) = E(z, ζ)dσg(z)
∂Ω ∂Ω

is valid in Ω for monogenic from the right (resp. left) functions f, g ∈ CΩ.
P3 ∂fk
5. Monogenic functions f = f (z) = 0 fk (x)ek with f0 = 0 and ∂x 0
= 0, k =
1, 2, 3 are monogenic from the left as well as from the right and describe an
irrotational fluid flow without sources nor sinks. This follows from the fact
that, in this case, the corresponding generalized Cauchy-Riemann systems are
equal, i.e. Df = f D = 0 and at the same time equivalent to the vector system:
→
−
div f = 0 (17)
→
−
curl f = 0.
where we identified in am obvious manner Imf = f = ∼− →T
f .

7.2 The Moisil-Teodorescu approach

Independently from Fueter, but also motivated by the idea to create a spatial
holomorphic function theory two Romanian mathematicians, G. C. Moisil and N.
Teodorescu, published in 1931 a paper [38] where they used a traceless matrix
differential operator of the form
 0 ∂ ∂ ∂ 
∂x1 ∂x2 ∂x3
∂
 ∂x 0 ∂
− ∂x ∂ 
D=
 ∂
1
∂
3 ∂x2
∂
.
 (18)
∂x2 ∂x3 0 − ∂x 1
∂ ∂ ∂
∂x3 − ∂x 2 ∂x1 0
3
Applying a matrix form of Stokes’ theorem in IR they studied the system
→
− −
→T
D f = 0 where f = (f0 , f1 , f2 , f3 ) (19)
which can also be written in the form
−
→
div f = 0 (20)
−
→
gradf0 + curl f = 0.

16
Comparing this with system (17) and its relation to monogenic functions one can
see that Moisil e Teodorescu studied a less restrictive case (f0 need not to vanish)
of the equation Df = 0, f : IH → IH.
Besides its historical value, this contribution it also has one feature which be-
came important later on as a general tool in Clifford Analysis: the fact that the
considered differential operator is not the generalized Cauchy-Riemann operator D,
but the Dirac operator

∂f ∂f ∂f
D= i+ j+ k. (21)
∂x1 ∂x2 ∂x3
It follows immediately that the Laplace operator in IR3 can now be factorized in
the simple form DD = DD = −D2 ∆3 which implies several advantages in the
applications to physical problems, where, from the viewpoint of spatial symmetry,
the choice of a distinguished variable like x0 is not motivated. Obviously, the use
of the hypercomplex derivative in form of Df also becomes senseless. The Dirac
operator is the central differential operator in fluid dynamics and in the theory
of heat conduction, but it also plays an important role in the description of the
electromagnetic field: Maxwell’s equations rely on the Dirac operator. But the
application from which the Dirac operator derives its name is quantum mechanics,
as it was mentioned before. In non-relativistic quantum mechanics usually the Dirac
operator on a three dimensional space is used as defined in (21). In relativistic
mechanics however one uses the Dirac operator on the Minkowski space IR1+3 .
This operator is no longer a linearization of the Laplacian ∆ = ∇2 , but of the
d’Alembertian , the operator of the wave equation.
From the mathematical point of view the use of the Dirac operator seems to be
particularly adequate for refinements in harmonic analysis and all problems were
only one differential operator is needed.

7.3 Trends in applied Clifford Analysis

About 75 years ago, in the beginning of Quaternionic analysis, the intention dom-
inated to create general methods for solving problems in other fields, for instance
in number theory and partial differential equations applied to special problems in
Mathematical Physics. Problems in dimensions higher than two have been the main
motivation for developing function theoretic tools in algebras more general than the
algebra of complex numbers. Soon it became obvious that the combination with the
algebra of quaternions or, more general, with Clifford algebras permits to develop
a whole theory, but only at the end of the 1960th started a systematical research
in this field.
In 1982, the book of Brackx, Delanghe and Sommen [4] marked the first period
of advances in theoretical research whereas the book of Gürlebeck and Sprössig
[20] already systematically described applications for solving linear and nonlin-
ear boundary value problems of the most important partial differential equations
of mathematical physics (Laplace and Helmholtz equations, equations of linear
elasticity, Maxwell equations, Navier-Stokes equations and others). The authors

17
studied questions of existence, uniqueness, regularity, and general representation
of their solutions in a unified form. Furthermore they introduced and developed
new boundary collocation methods as well as represented a discrete model of the
quaternionic function theory for constructing finite difference methods.
In 1992 Delanghe, Sommen and Soǔek [9] studied applications in quantum
physics, particularly in form of a function theory of the Dirac operator.
M. Mitrea [37] published in 1994 Lecture Notes dealing with Clifford wavelets,
thereby showing that this modern branch of applied mathematics can also take
profit from tools naturally adapted to higher dimensions.
Developing further ideas of [33] the application of integral representations for
spatial models of mathematical physics (e.g. general Helmholtz equation, electro-
dynamical models and massive spinor fields, including the MIT bag model in the
theory of quarks) was considered in 1996 in the book of Kravchenko and Shapiro
[33].
An almost complete picture of the state of the art in theory and applications of
Clifford analysis up to 1997 is contained in [21] which we mentioned already in the
first section.
Without being able to overview all recent trends in the vast research on appli-
cations of Clifford analysis it seems to us that those aforementioned ”traditional”
fields, mainly related to mathematical physics, are continuing and undergoing a
refinement. New results about Clifford-Hermite wavelets in Euclidean Spaces or
Cauchy Transforms on Rectifiable Surfaces which will be presented at this IKM are
two examples of such a development. Several other announced contributions join
this trend. Unfortunately it is not possible to mention all of them explicitly.
Of more theoretical nature are contributions in Clifford Analysis on Projective
Hyperbolic Space, but they are also aiming to future applications. It is our opinion
that they are the expression of an undergoing reinforcement in the fundamentals of
Clifford Analysis, which besides all progress are still not exhaustively constructed.
This concerns also topics in classical complex function theory which until now have
not yet been the target of Clifford Analysis research. As an example we refer to the
fact that a derivative concept is one of the main classical function theoretic concepts
for the qualitative and quantitative characterization of functions. Generalizations
in approximation theory, the characterization and solution of mapping problems on
higher dimensional manifolds or the qualitative investigation on scales of function
spaces are some of the fields where the hypercomplex derivative mentioned in section
7.1 has proved to be useful.
After 20 years of remarkable dynamics and success in several areas of theoretic
and applied research, Clifford Analysis will continue to approve its importance as
a combination of algebra, geometry and analysis, particularly adapted to higher
dimensions.

References

1. ABLAMOWICZ, R. FAUSER, B., RYAN, J., SPRÖSSIG, W.: Clifford Algebras and
their Applications in Mathematical Physics, Vol. 1 Algebra and Physics (AbÃlamowicz,
R., Fauser, B. (eds.) Vol. 2 Clifford Analysis, (Ryan, J., Sprößig, W.(eds.),

18
 Birkhäuser, 2000
2. ALTMANN, S. L.: Hamilton, Rodríguez, and the Quaternion Scandal (what is wrong
with on the major mathematical discoveries of the nineteenth century) Mathematical
Magazine, 62 ,1989, 291-308.
3. BAYLIS, W. E. : Electrodynamics: A Modern Geometric Approach, Birkhäuser, 1999
4. BRACKX, F., DELANGHE, R. , SOMMEN, F.: Clifford Analysis. Pitman 76,
Boston-London-Melbourne, 1982.
5. CLIFFORD, W.: Applications of Grassmann’Extensive Algebra, American Journal
of Mathematics Pure and Applied, 1, 1878, 350-358.
6. CORROCHANO, E. B., SOBCZYK, G. (eds.): Geometric Algebra with Applications
in Science and Engineering Birkhäuser, 2001.
7. CROWE, M. J.: A history of Vector Analysis, Dover, 1993.
8. DELANGHE, R.: On regular-analytic functions with values in a Clifford algebra,
Math. Ann. 185 (1970), 91-111.
9. DELANGHE, R., SOMMEN F., SOUCEK V. (Brackx F., Constales D.): Clifford
Algebra and Spinor- valued Functions, Kluwer Academic Publishers, 1992.
10. DIRAC, P. A. M.: The quantum theory of the electron, Proc. Roy. Soc. A117, 1928
11. Dorst, L., Doran, Ch., Lasenby, J. (eds.):Applications of Geometric Algebra in Com-
puter Science and Engineering, Birkhäuser, 2002.
12. DYSON, F.: Missed Opportunities, J.W.Gibbs Lectures, Bull. AMS, Vol. 78 (1972),
635-652.
13. FUETER, R.: Über Funktionen einer Quaternionenvariablen Atti Congr. Int. mat.,
Bologna 1928.
14. FUETER, R.: Über die analytische Darstellung der regulären Funktionen einer
Quaternionenvariablen, Comment. Math. Helv. 8 (1935-36), 371-378.
15. FUETER, R.: Über die Funktionentheorie in einer hyperkomplexen Algebra Ele-
mente der Mathematik, III, 5, 89-94, 1948.
16. FUETER, R.: Funktionentheorie im Hyperkomplexen, Vorlesung Wintersemester
1948/49 Ausgearbeitet und erweitert von Erwin Bareiss, 318 pp. Mathematisches
Institut der Universität Zürich, 1949.
17. GIBBS, J. W.: On multiple algebra, Vice-presidental Address to the Section of Math-
ematics and Astronomy of the American Association for the Advancement of Science.
Proc. Amer. Assoc. Adv. Sci., 1886, vol. 35, 37-66.
18. Grassman, H. G.: Die lineale Ausdehnungslehre - ein neuer Zweig der Mathematik,
Verlag Wiegand, Leipzig 1844
19. GÜRLEBECK, K., MALONEK, H.: A Hypercomplex Derivative of Monogenic Func-
tions in IRm+1 and its applications, Complex Variables 39 (1999), 199-228
20. GÜRLEBECK, K., SPRÖSSIG, W.: Quaternionic Analysis and Elliptic Boundary
Value Problems. Birkhäuser Verlag, Basel, 1990.
21. GÜRLEBECK, K., SPRÖSSIG, W.: Quaternionic and Clifford Calculus for Physi-
cists and Engineers, John Wiley &. Sons, 1997)
22. HAMILTON, W. R.: On a new Species of Imaginaries Quantities connected with a
theory of Quaternions - Proceedings of the Royal Irish Academy, Nov. 13, 1843, vol.
2, 424-434
23. HAMILTON, W. R.: Copy of a letter from Sir William R. Hamilton to John T.
Graves, Esq.,-Philosophical Magazine, 3rd series, 25 (1844), p.p. 489-95, edited by

19
 David R. Wilkins, 1999
24. HAMILTON, W. R.: On Quaternions, Proceedings of the Royal Irish Academy, Nov.
11, 1844, vol. 3 (1847), 1-16.
25. HAMILTON, R.W.: Elements of Quaternions, Longmans Green, 1866, reprinted by
Chelsea, New York, 1969.
26. HESTENES, D.: Space-Time Algebra, Gordon and Breach, N.Y., 1966.
27. HESTENES, D.: Multivector Calculus, J. Math. Anal. Appl. 24, 1968, 313-325.
28. HESTENES, D.: New Foundations for Classical Mechanics, D. Reidel Publishing
Company, 1986.
29. HESTENES, D.: Point groups and space groups, in: Applications of Geometric Alge-
bra in Computer Science and Engineering; Dorst, L., Doran, Ch., Lasenby, J. (eds.)
Birkhäuser, 2002.
30. HILBERT, D.: Die Theorie der algebraischen Zahlkörper Jahresbericht Deutsch.
Math. Verein.4, 175-546, 1897 in: Ges. Abh. I, Springer, 63 - 363, 1932.
31. HURWITZ, A.: Vorlesungen über die Zahlentheorie der Quaternionen, Springer,
Berlin, 1919.
32. IFTIMIE, V.: Fonctions hypercomplexes, Bull. Math. Soc. Sci. Math. Repub. Soc.
Roum., Nouv. Ser. 9 (57) (1965), 279-332
33. KRAVCHENKO, V., SHAPIRO, M.: Integral representations for spatial models of
mathematical physics, Addison Wesley Longman, Harlow, 1996.
34. KUIPERS, J. B.: Quaternions and Rotation Sequences, Princeton Uni. Press, 1999.
35. MALONEK, H.: A new hypercomplex structure of the Euclidean space IRm+1 and
the concept of hypercomplex differentiability, Complex Variables Theory Appl., 1990,
Vol. 14, 25-33.
36. MALONEK, H.: Power series representation for monogenic functions in IRm+1 based
on a permutational product, Complex Variables Theory Appl., 1990, Vol. 15, 181-
191.
37. MITREA, M.: Clifford Wavelets, Singular Integrals, and Hardy Spaces. Lecture
Notes in Mathematics 1575, Springer-Verlag, Berlin Heidelberg, 1994.
38. MOISIL, G.C., THEODORESCU, N.: Fonctions holomorphes dans l’espace, Bul.
Soc. Stiint. Cluij 6, (1931), 177-194
39. POMPEIU, M.: Sur une classe de fonctions dúne variable complexe. Rend. Circ.
Mat. Palermo, V. 33 (1912), 108 - 113, V. 35 (1913), 277 - 281.
40. SCHEFFERS, G.: Verallgemeinerung der Grundlagen der gewöhnlichen complexen
Zahlen, Berichte kgl. Sächs. Ges. der Wiss. 52 (1893), pp.60
41. SOMMRER, G.(ed.): Geometric Computing with Clifford Algebras - Theoretical
Foundations and Applications in Computer Vision and Robotics, Springer, 2001.

20