See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/355807649

When the mathematician David Hilbert tried in 1916 to build the first Theory
of Everything

Conference Paper · November 2021

CITATIONS READS

0 84

1 author:

Jean-Pierre Petit
Aix-Marseille Université
120 PUBLICATIONS 505 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Bimetric Cosmological Model View project

theoretical metaphysics View project

All content following this page was uploaded by Jean-Pierre Petit on 01 November 2021.

The user has requested enhancement of the downloaded file.

 1

When the mathematician David Hilbert tried in

1916 to build the first Theory of Everything
Jean-Pierre Petit1
Assisted by Pierre Marquet
for the translation from German

The present article is based on the two publications made by Hilbert in 1915
and 1916 ([1], [2]). Before entering into the heart of the matter, it is necessary to
specify the context in which these works were realized.

We are in 1915. At that time, only two forces at work in the universe were
known: the electromagnetic force and the force of gravity. The other two: the
strong interaction force, binding the components of the atomic nuclei, and the
weak force, responsible for beta radioactivity, would not be discovered until
much later.
The upheaval introduced by Albert Einstein, with his special relativity, was
finally accepted, at least by some advanced minds, since it is the only one to
account for the experiment initiated by the American Abraham Michelson in
1887, which concludes that the value of the speed of light is invariant, regardless
of the reference frame, fixed or immobile, in which we operate. No other credible
interpretation of the speed of light was found.
However, this idea took time to become one of the pillars of modern physics, so
much so that when the Nobel Prize was awarded in 1921 to its author, it was not
for this idea but for his interpretation of the photoelectric phenomenon. Einstein
is considered the inventor of the word "photon".

What is the discovery of relativity?

It is based on a new vision of the universe, with the appearance of a junction

between two words, leading to the compound word space-time. Einstein, thus,
can be considered as the inventor of space-time.
Previously, space and time were dissociated objects. Space is considered as
Euclidean. That is to say that the theorem of Pythagore in three dimensions, as if
we locate the position of two points A and B with the help of an orthonormal
frame of reference by giving them coordinates

{x A
, yA , z A } et {x B
, yB , z B }
the distance between them is:

1 Former Director of Research at CNRS (plasma physics, astrophysics, cosmology)

jppetit1937@yahoo.fr
 2

(x − xA ) + ( y B − yA ) + ( z B − z A )
2 2 2
L= B

Time is measured differently, with clocks, with mechanical systems. Before

Einstein, nobody would have had the idea to join, to mix two "objects" as different
as space and time, to combine meters and seconds.

Behind all this there is a geometrical vision of the cosmos. If we remove a

dimension of space, the z coordinate, for example, we have the following diagram:

Fig.1 : Pre-relativist space

As for the time t = 0, it obviously refers, in 1915, to the instant of the creation of
the universe, "by God". At that time, before the irruption of special relativity in
the mode of science and physics, the question "what is then the geometry of
space-time? It cannot be identified with a Riemanian mathematical space, defined
by a metric. Otherwise, what sense can be given to the following formula, defining
a "length » s :

(1) ds 2 = dx 2 + dy 2 + dz 2 + a 2 dt 2

where a would be a constant, in the form of a velocity, so that we can add up

similar quantities. This metric is then devoid of physical meaning.

The space coordinates are obviously real : { x, y, z } ∈! 3

What about time? It would never occur to anyone to imagine a negative time,
nor would it occur to anyone to imagine a retrochronic time flow. This variable t
therefore belongs to the set t ∈! + .

Albert Einstein's discovery has however a very clear geometrical interpretation,

through the space invented by the Russian mathematician Hermann Minkowski.
 3

Fig. 1 : Hermann Minkowski (1864-1909)

It is him who imposes this idea of a space-time continuum, defined by the way
in which expresses the length, according to a tool, qualified by the French
mathematician Henri Poincaré of pseudo-metric:

ds 2 = c 2 dt 2 − dx 2 − dy 2 − dz 2

The constraint, imposed in 1905 by Einstein's special relativity, is expressed in

a simple and clear way: it is enough to say that s is real. Thus it is necessary that:

dx 2 + dy 2 + dz 2
c 2 dt 2 − dx 2 − dy 2 − dz 2 ≥ 0 ou : v2 = 2
≤ c2
dt

This implies that the speed is less than c . To this metric, we associate its
signature, in the form of the sequence of its signs:
(+ − − − )

When Einstein undertook to describe gravitation using a bilinear form, if we

note by { x1 , x 2 , x 3 , x 4 } the coordinates of one of the points of the tangent space,
{ x ,x
1 2
, x 3 } locating it in space and x4 being the time coordinate, it seems logical
to write, indifferently :

ds 2 = dx4 2 − dx12 − dx2 2 − dx32 ou ds 2 = − dx12 − dx2 2 − dx32 + dx4 2 ≥ 0

This choice appears for example in the article published in 1915 by Einstein,
referring to his calculation of the advance of Mercury's perihelion [3] .
 4

Fig.2 : Einstein's signature choice

It is easy to see that this choice of signature is ubiquitous in the papers of all
authors who published papers prior to Hilbert's 1916 paper. Let us quote
Schwarzschild, Weyl, Droste, etc.

Some preliminaries before tackling Hilbert’s

conception of the geometry of space-time.
A 2D geometric object can be described by its metric. The metric of the sphere
is for example:

(1) ds 2 = R 2 ( dθ2 + sin 2θ dϕ 2 )

A metric that is totally regular, whatever the values of the two variables. How
can we say, on the basis of this simple fact, that it is a 2-sphere? Well, we perform
a change of coordinate θ = arcsin(r / R) to shift from the set { θ , ϕ } to the set
{ r , ϕ } . Then we get :
R2
(2) ds = 2 2 dr 2 + r 2dϕ
2

R −r

We then notice that for r = R the first term has a zero denominator. We have
thus created a coordinate singularity. Another remark: for r < R the term ds2 is
negative. The element of length ds becomes pure imaginary. This is normal: we
are outside the sphere. For these 2D metrics, defining objects, surfaces, we see a
thread emerging. The metric is a polynomial of degree two, a bilinear form,
expressed with a certain set of coordinates, a priori real. If the length element is
also real, it is because our definition interval has been judiciously chosen.
 5

Otherwise, where ds is imaginary, we are simply outside the surface. Of course, in

a formal way, we can always consider studying the behavior of this object,
outside this interval of definition. But then we break our rule of the game. We are
no longer in the world of the real, but in the world of the complex.
Let's go back to the question we asked. How do we know that (1) and (2)
represent a sphere? To do this, we will plunge it into a three-dimensional space
{ r , ϕ , z } . The physicist recognizes the "cylindrical" coordinates of a 3D
Euclidean space. In this imbedding operation the length element must be
expressed in the same way, especially on meridian curves with constant . We will
therefore write:

R2
(2) ds = 2 2 dr 2 = dr 2 + dz 2
2

R −r

It is a differential equation which immediately gives us the link between r and z.

Its integration gives us:

(3) r 2 + z 2 = R2

This surface is thus generated by the rotation of a circle centered at the origin
of the coordinates, around the oz axis. It is indeed a sphere S2. We could do the
same thing starting from two expressions of the metric of the torus T2:

dr 2
(4) ds 2 = rg 2 dθ2 + (Rr + rgcosθ )2 dϕ 2 et ds 2 = + r 2 dϕ 2
− r 2 + 2 r Rr + rg 2 − Rr 2

In these expressions we recognize the radius rg of the generating circle of the

torus, the radius of this small circle whose center turns around an axis passing by
its plane, along a circle of radius Rr. On the left we have opted for coordinates
{ θ , ϕ } which are not a problem. On the right, switching to the representation
system { r , ϕ } we have, as for the sphere, created coordinate singularities for
the two values that cancel the denominator of the first term of the second :

(5) r = R r + rg et r = R r - rg
Moreover the ds is real only if this denominator remains positive:
(6) R r - rg < r < R r + rg
otherwise we are outside the surface. An imbedding operation in the three-
dimensional Euclidean space allows to discover geometrical properties which
will make appear the mode of generation of the torus. But nobody is interested in
the geometrical properties of this object, for example for r < (Rr - rg) . If we
decided to do so, we would leave the mode of the real to enter a strange complex
geometry, which then has nothing to do with tangible 2D objects.
These are objects defined by metrics whose signs are all positive. We will call
them elliptic metrics. This is true for an unlimited number of dimensions, let's
consider for example the 3D object:
 6

(7) ds 2 = dr 2 + r 2 ( dθ2 + sin 2θ dϕ 2 )

There, using adequate coordinate changes to make the metric of a Euclidean

space reappear:

(8) ds 2 = dx 2 + dy 2 + dz 2

Thus we would find our familiar representation space. In (7) the points are only
marked in 3D polar coordinates. But, by studying the object with the help of a
family of surfaces with constant r, fitting together like Russian dolls, we could
"read" the object with the help of this folding method. This space has geodesics
which are the infinite number of lines that we can draw in this 3D space. Among
these are the straight lines coming from the origin, which are perpendicular to
these surfaces S2 which are spheres. The coordinates { r , θ , ϕ } are Gaussian
coordinates, a concept to which Hilbert will refer in what follows.
We can leave the 3D Euclidean by imagining 3D spaces, 3D hypersurfaces defined
by:

(9) ds 2 = f (r) dr 2 + r 2 ( dθ2 + sin 2θ dϕ 2 )

and we would have the same folding system by spheres. We will consider the
particular case:

dr 2
(10) ds =
2
+ r 2 ( dθ2 + sin 2θ dϕ 2 ) Rs > 0
R
1− s
r
For the moment there is no time variable. In the perspective of an extension of
what we said about 2D surfaces to 3D hypersurfaces, defined by their metric,
from which we can build their geodesic curves, we will consider the existence of
this hypersurface in its ds is real, so when ds2 > 0. This gives a defining space such
that r > Rs . By making r constant we can still leaf through the object with a family
of spheres nested one inside the other like Russian dolls. But there is then a
sphere of minimal area 4π Rs 2 which corresponds to the metric:

(11) ds 2 = Rs 2 ( dθ2 + sin 2θ dϕ 2 )

It is a throat sphere. But what happens in r = Rs? The denominator of the first
term of the second member becomes zero. Is this sphere singular? No, it is still a
coordinate singularity. We can eliminate it by changing the variable:

(12) r = Rs ( 1 + L n chρ )

The metric thus becomes:

1 + L n ch ρ 2
th ρ d ρ 2 + Rs2 ( 1 + L n ch ρ ) ( dθ )
2
(13) ds 2 = Rs2 2
+ sin 2 θ dϕ 2
L n ch ρ
 7

There is then no limit in the definition space, and can vary from minus infinity
to plus infinity. The metric potentials are:
(14)

1 + L n ch ρ 2
g ρ ρ = Rs2 th ρ
L n ch ρ
gθθ = Rs2 ( 1 + L n ch ρ )
2

( 1 + L n ch ρ )
2
gϕϕ = Rs2 sin 2 θ

How did this operation make this metric regular? When the hyperbolic cosine
is unity and its logarithm is then zero. So the denominator in the first term of the
second member is always zero. Yes, but the same is true for the hyperbolic
tangent. If we do a series development in the neighborhood of ρ = 0 you will see
that g ρ ρ → 2 . This hypersurface is therefore perfectly regular. Its throat sphere
for ρ = 0 has a minimum area equal to 4π Rs 2 . To calculate this "2D volume" (a
surface) you must do:

(15) ∫∫ g dθ dϕ

But you can convince yourself that the non-contractibility of the object is still
present. All you have to do on this sphere is to make θ = π / 2 and to vary ϕ from
0 à 2π . You get a finite perimeter p = 2π R s . What happened? You are no
longer in your comfortable three-dimensional Euclidean representation space
(the only one you have in fact, to build a mental image).
This hypersurface is therefore a three dimensional manifold, equipped with an
elliptic Riemanian metric. In this new system of axes the determinant is never
zero. This means that this hypersurface is orientable. At any point one can define
a vector product and the "corkscrew rule", which goes with it, will be the same at
all points. It is obviously painful for the neurons to consider this kind of "space
bridge" which creates a passage between two Euclidean 3D spaces (which are like
"one inside the other"). We can call it a "3D diabolo". We will see later how the
mathematician Hermann Weyl created and studied this object, in 1917.

By the way, you can take these steps with the "2D diabolo", defined by the metric:

dr 2
(16) ds 2 = + r 2 dϕ 2 Rs > 0
Rs
1−
r
With the same change of coordinate you could check its regularity by obtaining:

1 + Logch ρ 2
th ρ d ρ 2 + Rs2 ( 1 + Logch ρ )
2
(17) ds 2 = Rs2 dϕ 2
Logch ρ
 8

But then there is a much more "tangible" way to apprehend this surface. It is
enough to imbed it in a three-dimensional space and to build its meridian
( ϕ = cst ) . You then obtain, as for the sphere, the differential equation:

dr 2
(18) ds 2 = = dr 2 + dz 2
R
1− s
r
Its solution is the "lying parabola":

z2
(19) r = Rs +
4Rs

This is the surface:

Fig.3 : The 2D diabolo.

We will find this pattern in the analysis made in 1917 by Weyl, which we will
detail later.

Another system of representation: projection, and the traps of thought.

We have evoked a mode of representation of a 2-surface by imbedding it into

our 3D Euclidean space. We make there a transcendent gesture, by adding an
additional dimension. But how a being living in a 2D euclidean space would
represent this diabolo? He could only conceive it projected in his own world. He
would then imagine a strange border, represented by a circle. The objects which
go on this surface, not Euclidean, cross then a circle of throat. Our inhabitant of
the Euclidean space 2D can then imagine that his mode "has a place and a
reverse ».
 9

Dig.4 : Plane representation of the 2D diabolo

We can illustrate this relation of enantiomorphy by starting from an oriented

triangle drawn on this plane, "habitat of our 2D observer". The figure below
illustrates this inversion of the orientation.

Fig.5 : Reverse orientation.

 10

This seems obvious to us because we have the possibility of imbedding these

two structures into our 3D Euclidean representation space. But for the inhabitant
of this plane it would be very problematic to consider "that inside this circle there
is nothing".

Let's move on to the 3D hypersurface. Here, we can no longer draw, but the
idea of projection into a 3D Euclidean representation space is the same. This time,
the inhabitant of this space is you, it is me. It will be very difficult to consider
"that inside this throat sphere there is nothing", and that one cannot contract a
sphere by giving it an area lower than a finite value, in a word that this 3D space
is not contractible.
Through these 2D and 3D examples we see that the fact of using a Euclidean
space of representation (the only mental tool we have) to try to read, to
"understand" (etymologically "to take together") objects presenting themselves
in the form of sets of points leads us to imagine objects that, in fact, do not exist.
This is particularly striking for the 3D structure where we are totally unable,
mentally, to get rid of this idea of "the inside of the throat sphere".

Hyperbolic surfaces and hypersurfaces.

The word hypersurface always evokes a possible representation in a higher
dimensional representation space. We have an intuitive image of the geodesics of a
surface. It is much more difficult to imagine them in 3D. In general relativity it is
often said that the space-time is a hypersurface with four dimensions. Here again,
the object is defined by its metric. What Einstein and Minkowski have brought is
the introduction in physics of hyperbolic metrics, whose signature makes
opposite signs cohabit. We can thus consider a relativistic space-time with two
dimensions

(20) ds 2 = c 2 dt 2 − dx 2

The calculation of geodesics corresponds to the variational problem:

(21) δ ∫ ds = 0
AB

We look for curves corresponding to paths where the distance traveled is

minimal. This leads us to solve the Lagrange equations. These lead us to
representations of x and t that are linear as a function of the parameter s.
Consequently x and s are linked by the linear relation x = v s, where v is the speed.
And if we impose that the length s is real, we must have:

dx
(22) v= < c
dt

What do the theorists do then? What do we find in all the books, the courses?
We find images like these, in 2D or 3D:
 11

Fig.6 : The light cone.

On the left a way to represent this hyperbolic space (t,x) in two dimensions. On
the right the classical "light cone". On the left figure the red curve is supposed to
represent a path corresponding to x = v t with v < c
The black curve represents a path with v > c . It is located in the "elsewhere".
But in doing so, what are we doing? We are trying to build a 2D image of a
hyperbolic space by projecting it into a 2D Euclidean space, of metric:

(23) ds 2 = c 2 dt 2 + dx 2

We thus create "something that does not exist", in this case this greyed-out
surface or volume "outside the light cone". This space does not exist any more
than this "interior of the throat sphere" that we create by trying to project the
structure of the 3D diabolo into a 3D Euclidean space. This "elsewhere" exists,
etymologically speaking, only in our imagination and stems from the image we
have created.

The conclusion is simple:

This preamble having been made, we will move on to the subject of the article,
to the way Hilbert created his own representation of space-time and hence of the
universe.
 12

The hectic race, neck and neck, of two geniuses.

In 1915 Hilbert was 53 years old and already had an impressive record of
achievement behind him, which had made him known far beyond the German
borders. He loved abstraction and logic and was known, among other things, for
publishing a treatise in which he defined the axioms underlying Euclidean
geometry. All German and foreign mathematicians consider him a "beacon" in
the discipline and know that his name will go down in the history of mathematics.
He was not always interested in physics. An amusing anecdote is reported about
him. When he was asked to replace the mathematician Felix Klein, who every
year gave a lecture to the students of an engineering school in Göttingen, he
began his lecture with these words:

- It is said that mathematicians and engineers have difficulty understanding

each other. This is not true: they simply have nothing to do with each other.

Fig.7 : David Hilbert (1862-1943), in 1915

 13

Fig.8 : Félix Klein ( 1849-1925) in 1915

It should be noted that at that time a similar gap existed in the field of
experimental physics, at its fundamental level, where people who were physicists
and chemists were working on what would later be called nuclear physics.
Among them was the colorful New Zealand physicist Ernst Rutherford. Solicited
during the First World War by politicians, who asked him if he could not produce,
from his work, some new weapon that would allow England to overtake its
adversary, Germany, he had answered them, as he was laying the foundations of
the future nuclear physics:

- I leave it to your chemists to invent asphyxiating gases and to your

engineers to invent planes, submarines and torpedoes. We, scientists, are
concerned with totally different things, seeking to penetrate the secrets of
matter.
It was the meeting with the young Einstein, twenty years his junior, that was
decisive for Hilbert. He then discovered a fantastic field of applications of
sophisticated mathematics to physics, which from then on was no less
sophisticated. He established close relations with Einstein, which could even be
described as friendly and which were in any case based on a great mutual esteem.
In June 1915 Einstein gave him a real lecture on relativity and Hilbert understood
that there was a way to use the extremely powerful tool represented by the
techniques of calculus of variations.
He began by applying this idea to electromagnetism, and then he became
interested in gravitation. At the time, the telephone did not exist. It is thus
through numerous letters that these two communicate. It so happens that a
significant number of these correspondences have come down to us. They are
reproduced, in whole or in part, among others by Tilman Sauer [3,], who has
kindly reproduced these texts in their English translation. Hilbert communicated
 14

to Einstein his vision of things: he thought he was on the verge of unifying the
only two forces known at that time, the electromagnetic force and gravitation.
Einstein worked differently, by trial and error. Great mathematics is not his forte.
Very intuitive, being above all a fantastic physicist, it is by trial and error that he
is about to arrive, after ten years of reflection, at what will be considered as the
key to a new theory, that of general relativity. But Hilbert beat him to the punch
on November 20, 1915 [1]. It was only five days later that Einstein sent to the
same journal: the Annals of the Prussian Academy of Sciences [4]:

Fig.9 : Albert Einstein's November 25, 2015 article entitled:

"The Field Equation of Gravitation." [4]

Hereafter is Einstein's equation where he formulates what both of them were

chasing, an equation whose two members are at zero divergence.
 15

Fig.9 : The Einstein field equation in its first form (25 nov. 1915).

Alas, four days earlier Hilbert wrote in his article this equation:

Fig.10 : The Hilbert field equation (20 nov. 1015) [1].

 16

Translation :

Using the notations of the variational derivation with respect to that we

introduced above, the equation of gravitation takes the form:

⎡ g K⎤ + ∂ g L = 0
⎣ ⎦µ ν ∂g µ ν

With, as first member

⎡ g K⎤ = g ( K − 1 K g )
⎣ ⎦µ ν µν
2
µν

For Hilbert K µ ν is the Ricci tensor, which Einstein calls Rµ ν . K is the scalar
derived from it, the "Ricci scalar", which is designated by the letter R in Einstein.
Moreover, the:

∂ gL
= − g Tµ ν
∂g µ ν

So the Hilbert equation is written:

1
(24) Rµ ν − R gµ ν = Tµ ν
2

It is indeed in this form, which lacks the term attached to the cosmological
constant:

Λ gµ ν

that Einstein will introduce later, on Hilbert's advice, to succeed in building the
first relativistic cosmological model, describing a stationary universe, that this
equation will enter history.
The equation in Einstein's paper, which is 5 days later than Hilbert's, is just
another equivalent form, where we can recognize, on the right, "the matter tensor
Tim » and « the Laue scalar» T, which derives from it 2 . But Hilbert adds a
construction of the equation by variational method by basing it on an action built
on what he calls a "function of the universe":
(25) H=K+L

2 Built from the matter tensor, in the same way that the Ricci scalar R derives
from the Ricci tensor Rim.
 17

K is obviously the Ricci scalar. This technique will be remembered in history as

the "Einstein-Hilbert action".
In the course of their intense correspondence, very warm relations were
created between these two geniuses, who assumed themselves as such, perfectly
conscious of their own value, and that of their partner.

Fig.11 : Albert Einstein in 1915

Indeed, Einstein and Hilbert were pursuing, at a time described as "hectic", two
parallel research programs, where Einstein focused on gravitation alone, while
Hilbert dreamed of uniting the two force fields, electromagnetic and gravitational.
In [3] you can read about these days of agitation and doubt. But finally Einstein
chose to calm down, writing to Hilbert on December 20, 1915 (see this reference,
page 48) :
- There was a certain resentment between us, the cause of which I do not
wish to analyze. I have fought against the bitterness associated with it,
and with complete success, I think of you again with unmixed friendship,
and I wish you would think the same way.
In all rigor the conclusion should have been to attribute this equation to both
authors. But at that time Einstein's steps were not yet completely assured. On the
other hand, Hilbert's brilliant career is a recognized fact. The elder one, very
sporty, leaves the premium to the younger one.

However, when we go back to these two equations there is a difference. In

Einstein's equation there is a radical − g where the determinant of the metric is
preceded by a minus sign, absent in Hilbert. In what follows we will discover why.
 18

The cosmos in 1915.

Before presenting David Hilbert's own vision, it is important to consider what

scientists knew about the universe in 1915.

Spectroscopy was born in Germany, in Heidelberg, in 1859 with the first works
of Gustav Kirchhoff and Robert Bunsen, inventor of the gas burner of the same
name. It allows to identify the nature of a source on the basis of its spectral
signature. At the same time, with the help of a telescope installed on the roof of
the Vatican, Father Angelo Secchi pursued the idea that each star is linked to a
deceased person. But in the United States, in New England, at the same time the
astronomer Edward Pickering undertook a vast classification of the stars,
according to their spectrum, which will give birth to the Hertzprung-Russel
diagram in 1900. In 1865 the Scottish genius James Clerk Maxwell published the
equations that govern electromagnetism.

James Clerk Maxwell 1831-1879

(died at 48 years old)

In 1916, the Englishman Eddington followed very closely both the experimental
and observational advances and the progress of the theory. He participated
closely in the movement that led to the understanding of the mechanisms of
energy production in the form of radiation within stars, which would exploit the
understanding of radioactivity, a theory that would only become functional in
1920.
Since 1840 we know the velocities of stars (and beyond their masses),
evaluated by the effect discovered by the Austrian Christian Doppler (1803-
1853) and the French Armand Fizeau (1819-1896).

Thanks to Newton, celestial mechanics had taken its bearings. Confronted with
the problem of the instability of the trajectories of the planets, he believed that,
 19

from time to time, it was God who, operating behind the scenes, put them back
into their orbits. This idea was invalidated by the Frenchman Pierre Simon de
Laplace who solved the problem using mathematics. Questioned by Bonaparte,
who asked him what was the place of God in all this, he replied "that he did not
need this hypothesis in his calculations". In 1902, the Englishman James Jeans
formalized the mechanism of gravitational instability giving rise to stars and
planets.
The first elementary particle discovered was the electron, by the Englishman
J.J.Thomson in 1897, thus at the immediate dawn of the century. The idea results
from the interpretation of the experiments carried out by the Englishman
Crookes where a cathode placed in a vacuum tube projects its "cathodic
radiation", which is deflected by a magnetic field. In 1995 the Frenchman Jean
Perrin identified these "rays" as a jet of electrons. Very quickly the ratio mass-
charge and determined and the latter is measured in 1911 by the American
Millikan.

Fig.12 : J.J.Thomson 1856-1940

When in 1905 the New Zealander Ernst Rutherford demonstrated the

corpuscular nature of matter and the existence of atoms, the idea of atoms made
up of positively charged nuclei around which electrons gravitate emerged in
1913. The first model, which proposed to describe the hydrogen atom, had been
put forward by the Dane Niels Bohr, who was then 28 years old.
Thus, during the first half of the 19th century, all the tools, both theoretical and
observational, were developed. When Hilbert wondered about the functioning of
the cosmos, the quantity of discoveries made in the few preceding years was
mind-boggling and contrasted with the current stagnation of physics, astronomy
and astrophysics, for half a century, where no new particle has been discovered,
where the "scientists", the "finders" seem to have been replaced by an army of
"researchers", five hundred times more numerous than their elders,
"accumulating the data".
 20

It remains that nobody imagines for a single second any evolution of the cosmic
scene, perceived as globally homogeneous and stationary. The idea of a creation
by God, at an "instant zero", is imposed in everyone's mind, whether explicitly
formulated or not.
Let us add that the very beginning of the discovery of the deep nature of the
force of gravity, reinterpreted in terms of geodesics of a very weakly curved
space, does not change the global nature of the cosmos.
In this year 1915, which was decidedly rich in scientific events of the first
magnitude, we saw that Einstein had published, on November 25, 1915, a paper
presenting the equation of the gravitational field [4]. But, on the same day, he
brings a second important contribution in the form of a first linear solution,
which gives a precise evaluation of the advance of Mercury's perihelion [5]. The
linearization is amply justified, the phenomenon, minimal, a few tens of seconds
per century, can be assimilated to a disturbance. This result sounds like a
thunderclap. Not only the approach of the sky phenomena with the help of a field
equation represents a major paradigmatic leap, but a solution of this same
equation brings the key of an enigma remained until now without solution. A
work which is confirmed a few weeks later when the Austrian Karl Schwarzschild
[6] , who writes to Einstein, in December 2015, that he has just constructed the
linear solution.

This stimulated Hilbert who, throughout 1916, feverishly prepared the

publication of an even more ambitious work, an extension of his November 1915
paper, which he published on December 23, 1916, on Christmas Eve. At the same
time he announced to Einstein that he was about to synthesize electromagnetism
and gravitation, in short to create what can be considered as the first Theory of
Everything.

Hilbert's conception of the geometry of space-time.

If we except these tiny curvatures linked to the presence of masses, space-time

remains almost flat, almost Euclidean. Einstein has, of course, brought the idea
that motion alters the flow of time, but the phenomenon seems to manifest itself
in a sensitive way only for "relativistic" speeds, not negligible in front of the
speed of light, which do not yet belong to the world of experimental physics and
in any case totally negligible in Nature. In the following figure, we can give
Hilbert's representation of space-time.

Of course, the general formalism of relativity shows that the points of this space
can be located by an infinite number of different coordinate systems, just as there
is an infinite number of ways to locate the points on a sphere, to map it. But
Hilbert keeps in mind that there must be a particular system, better than the
others, that he calls "true". (Eigentliche3). The underlying idea is that of a 4D
space layered by an infinity of 3D hypersurfaces stacked on top of each other.

3 That which is most suitable, most appropriate, most in tune with reality.
 21

Fig.13 : The Hilbert space-time, in its primitive form.

The mass points obviously follow geodesics of this time space. They go then
according to "time lines" (Zeitlinie). The drawing shows this family of time lines,
perpendicular to the 3D hypersurfaces representing space. Of course, if there
were no force, the 3D hypersurfaces of space would be 3D Euclidean spaces, and
the trajectories would be parallels, perpendicular to these parallel folds. And
nothing would happen. But the gravity forces result in a barely perceptible
warping of the 3D hypersurfaces. Correlatively, the time lines deviate slightly
from this family of parallel lines. This being the case, we remain very close to a
Euclidean structure. Hilbert therefore starts by introducing four coordinates
( w1 , w2 , w3 , w4 ) which he designates as "universal parameters" (Weltparameter).
The first three coordinates refer to space, while w4 refers to time. The instant
w4 = 0 represents, even if Hilbert does not formulate it explicitly, "the instant of
the creation of the world by God" (Hilbert comes from a protestant family, deeply
religious).

For Hilbert, time is of a different essence than space. The elegant way to
account for this is to imagine that the coordinate of time is purely imaginary,
which he does by writing, on the first page of his paper :

(26) w1 = x1 w2 = x2 w3 = x3 w4 = ix4
 22

Before talking about length, as a good mathematician, Hilbert considers a

bilinear form G constructed from the coordinates ( X 1 , X 2 , X 3 , X 4 ) of a vector of
this four-dimensional space :

G ( X 1 , X 2 , X 3 , X 4 ) = ∑ gµ ν X µ X ν
µν

He does not specify his choice concerning this form until six pages later:

Fig.14 : The bilinear form preferred by Hilbert,

expressed in a Gaussian coordinate system.

This expression can then be expressed in its differential form:

(28) G ( dx1 , dx2 , dx3 , dx4 ) = ∑ gµ ν dxµ dxν

µν

Hilbert then considers a curve whose points are marked with a parameter p :

(29) xs = xs ( p) ( s = 1, 2 , 3, 4 )

It specifies well that these coordinates the xs ( p) are real. He then divides his
curve into portions and considers the expression :

dx1 dx2 dx3 dx4

(30) G( , , , )
dp dp dp dp
 23

He then considers two cases.

Either the bilinear form, in the region where this curve is spanned, is positive:

dxs
(31) G( )>0
dp

He then decided to call these portions of the curve segments (Strecke). He then
introduced a first "length of this segment", according to:

dxs
(32) λ = ∫ G( ) dp
dp

Then he considers another portion that he decides to call a timeline (Zeitlinie)

where :

dxs
(33) G( )<0
dp

and the integral calculated along this other portion of the curve will be called the
proper time of this timelike curve:

dxs
(34) τ=∫ − G( ) dp
dp

Finally it introduces curves of zero length (Nullinie), such as:

dxs
(35) G( ) = 0.
dp

Thanks to the expression (34) Hilbert manages to establish a link with

Einstein's relativity. Indeed its bilinear form, in its differential form, is:

(36) G ( dxs ) = dx12 + dx2 2 + dx32 − dx4 2

It is therefore negative if we want to fit with the requirements of relativity, with

a speed of light c = 1 :

dx12 + dx2 2 + dx32

(37) v2 = <1
dx4 2

Thanks to the introduction of the minus sign under the root, Hilbert finds the
proper time of special relativity:

(38) dτ 2 = dx4 2 − dx12 − dx2 2 − dx32

 24

But now you have the origin of the change of signature, which has imposed
itself as a standard today, in cosmology as well as in theoretical physics, without
finding a trace of an article that justifies it.

In the expression (37) we find the signature introduced by Einstein which is

( + − − − ) . In what Hilbert writes it has become ( + + + − ) or ( − + + + ) and the
signs were reversed.
In these first pages of his article of 1916 we find something much more serious,
which will weigh on all the later development of cosmology, which is this idea of
endowing its four-dimensional variety M4 , not with one length, but with two, λ
and τ ! He thus speaks of two different measuring instruments. The length will be
measured with a "light clock" (Lichtuhr). In other regions of his space-time,
where his G-form is positive, he will use a tape (Maβfaden ) to measure his length
λ . But he points out that if you try to measure in one region with the instrument
used in the other, it does not work. One obtains indeed then imaginary values.

He will not say more, in the rest of the article, about the nature of these
mysterious regions of his space-time where the lengths are measured with this
scalar. And this while he concentrates with great insistence on what can have,
according to him, a physical meaning (Physicalischer Natur).
A little further on, Hilbert defines the light cone (Null-kegel: the null cone)
which is in as = ( x1 , x2 , x3 , x4 ) (its vertex) and whose current point has
coordinates ( X 1 , X 2 , X 3 , X 4 ) satisfying the equation:

(39) G ( X 1 − x1 , X 2 − x2 , X 3 − x3 , X 4 − x4 ) = 0

And he specifies that all the time lines (timelike curves) resulting from the point
as are located inside this four-dimensional part of the world whose border is the
temporal separation of as .

All time lines (time-like curves) from the point as are located inside this four-
dimensional part of the world whose border is the temporal separation of as . He
then focuses on the problem of causality in physics, looking for a "true"
coordinate system.
On this subject he states :
- A space-time coordinate system is called "true" (Eigentliches Raum-
Zeitkoordinatensystem) if it is a system for which the following four
inequalities are satisfied, with the additional condition that the determinant
is negative.
 25

(40)

He then adopts the definition of a system of a change of space-time coordinates

also called "true", "real" (eigentliches). It is simply the system of coordinate
changes which, to "true" coordinates, satisfying the inequalities (40), makes
another system correspond, endowed with the same properties. The four
inequalities mean that at any point event as the associated null cone excludes the
linear space x4 = a4 but contains inside the line through the point
( x1 = a1 , x2 = a2 , x3 = a3 , x4 = a4 ).

He then considers a line of universe xs = xs(p). From (33) it follows that in a

"true" space-time coordinate system we must have :

dx4
(41) ≠0
dp

He deduces that along this time line, the "true" time coordinate x4 must be
systematically increasing and cannot decrease. Because a time line remains a
timeline under any transformation of the coordinates, two point events located
on the same time line can never correspond to the same value x4 of the time
coordinate, through a "true" space-time transformation. This means that these
two events cannot be simultaneous.
We thus see that the cause and effect relations underlying the principle of
causality (Kausalitätsprinzips), do not lead to internal contradictions in this new
physics, if we take into account the inequalities (31) as part of our basic
equations, which leads us to confine ourselves in "true" space-time coordinates.
Hilbert now introduces the important point of his presentation: the use of
coordinates that he decides to call Gaussian, because they represent a
generalization of the polar coordinate system used by Gauss in his theory of
surfaces. Figure (13) illustrates the concept, which is a foliation, where these
surfaces correspond to a constant value of x4. The family of orthogonal curves are
geodesics along which this coordinate runs. If we opt for a coordinate system
where g 44 = − 1 then the x4 coordinate is identified with the proper time . The
Gaussian coordinates then satisfy the relation (32) and correspond to:

(42) g14 = 0, g 24 = 0, g 34 = 0, g 44 = − 1 4

Mais, à ce stade, le choix de Hilbert reste arbitraire. Il avoue échouer à le faire

émerger sur la base de considérations purement mathématiques.

4 In the translation published by Springer an error has crept in. We find g 44 = 0

 26

How Hilbert justifies his choice.

When a Riemann space is put in its diagonalized form (absence of crossed

terms) the sequence of signs attached to the different terms represents its
signature. It is an invariant by change of coordinates (real). Thus the Einstein
space-time corresponds to the bilinear form:

(43) G(dxs ) = g 44 dx4 2 − g11 dx12 − g 22 dx2 2 − g 44 dx332

and his signature is ( + − − − ) . In contrast, in his vision of space-time the form

retained by Hilbert is :

(45) G(dxs ) = − g 44 dx4 2 + g11 dx12 + g 22 dx2 2 + g 44 dx332

and his signature is ( − + + + ) . Einstein's choice is based on the identification of

the fourth coordinate through x4 = ct . Thus, the simple fact of imposing that the
shape is positive translates this physical property which is the limitation of the
speed to the speed c, that of light (corresponding to a zero value of the line
element). The proper time is then calculated by :

dxs
(46) τ = ∫ G( ) dp
dp

Hilbert's choice is less clear and forces him to situate the real, time-like
trajectories, in a representation such that G < 0. He must then introduce a change
of sign, and opt for the relatio:

dxs
(47) τ = ∫ − G( ) dp
dp

To this we can add the idea, singular, of endowing space with a second
measurement tool, giving a length, whose physical nature is not even touched
upon in the article, nor in the following ones, and which will give birth to the
myth of "spacelike curves".

It is then necessary to understand why Hilbert made this choice.

The explanation of the choice of an inverted signature by Hilbert.

For Einstein, this fourth coordinate is "of the same nature" as the other three.
Let's take our familiar three-dimensional Euclidean space. On the surface of the
Earth we will speak for example of the "length x", "width y" and "height z" of an
object, knowing that these denominations are arbitrary. We can make an element
of the group of rotations act on this object by totally modifying this scheme, while
we do not alter the object itself. The distances between its different points remain
unchanged. These rotations are part of the isometry group of this 3D Euclidean
 27

space. Instead of rotating the object we can decide to observe it from a different
angle.
The objects of the space-time are movements, characterized by the energy E
and the impulse p which are associated to them 5 . The isometry group of
Minkowski space, the Poincaré group, has a subgroup, the Lorentz group, which
happens to be the equivalent of the group of rotations and symmetries, in the
Euclid group. The latter operates rotations and symmetries that preserve lengths
(isometry: same length). The Lorentz group "operates rotations in four
dimensions" and preserves a length, that of the impulse-energy quadrivector.
These who see things in this way, through this "group-view" are then tempted
to imagine that this x4 dimension is real and is measured in ... meters. What
creates these "hyperbolic rotations" then comes from the axiomatic construction
of the Lorentz group:

⎛ 1 0 0 0 ⎞
⎜ ⎟
0 −1 0 0
(48) LT G L = G with the Gramm matrix G =⎜ ⎟
⎜ 0 0 −1 0 ⎟
⎜⎝ 0 0 0 −1 ⎟⎠

This aspect is present in Einstein's thought, but not in Hilbert's, who is

immediately attached to the reassuring Gaussian coordinates. Certainly, the mass
points do not remain immobile. But in a world which remains very far from
relativistic physics, we are very close to the Euclidean vision evoked in figure 13.
So what is real for Hilbert are the space coordinates :

( x1 , x2 , x3 )

These are tangible for him. Time is another matter. Nobody can take a second
between thumb and forefinger. So, Hilbert concludes, this one must be of another
nature, imaginary. Finally, another point, the universe has a beginning, in x4 = 0.

Hilbert is protestant. One can imagine him paraphrasing Genesis:

- God first created a four-dimensional space whose points were

marked by the coordinates
( w1 , w2 , w3 , w4 )

Before the play began, the author had to set the stage before the curtain rose.
All the objects were put in their place, the stars and their procession of planets,
ready to launch themselves into their orbits. The distances that separated them
being predefined and real.

5 The mathematician J.M.Souriau added the spin, as an object of pure geometry.

 28

- Then God decided that the fourth coordinate should be purely

imaginary, according to w4 = i x4 and that was the first day, the first
moment. The Earth then began its movement around the Sun, the
same as the other stars of the cosmos. Time appeared, irreversible
and implacable.
We read this interpretation of Hilbert's thought through these lines of his article.
Let us quote him:
Due to the nature of the three-dimensional space in x4 = 0 , as we have
posed it a priori (Vorausgesestzen: "posed first, before") the quadratic form
of the variables X1 , X2 , X2 described by the right-hand side of (see Figure
14) is necessarily positive and defined.
Therefore the first three inequalities (40 in this paper) are satisfied, as
well as the fourth, and the Gaussian coordinate system appears to be the
true (eigentliches: proper) coordinate system of spacetime.
What creates events? Hilbert takes up the idea launched a century earlier by
Laplace. He writes:
- If in the present time we have the data concerning the physical
quantities and their first derivatives with respect to time, their future
values can always be determined: the laws of physics, and this without
exception, the laws of physics, have been expressed to date through a
system of differential equations in which the number of unknown
functions is equal to the number of these equations.

Hilbert lists these data. These are the ten potentials gµ ν (µ ,ν = 1,2,3) which
emerge from the symmetrical tensor of format (4,4), representing the matter. To
this we must add, at any point, the components of the quadrivector
qs (s = 1,2,3,4) of electromagnetism. This makes a total of fourteen potentials.
But, when counting the equations, including those found in 1840 by the genius
Maxwell, Hilbert counts only ten, independent of each other.
Under such conditions, as is the case in this new physics of general relativity, he
concludes that it is not possible, from the knowledge of physical quantities at the
present moment, to determine future values in a unique way. Faced with this
impossibility of anchoring a logic of physics on a causality based on concrete
elements, Hilbert falls back on the opinion that "to follow the essence of this new
principle of relativity one must require the invariance, separately, of each postulate
of physics that has a physical meaning". And he adds "In physics, we must consider
everything that is not invariant by change of the system of coordinates, as devoid of
physical meaning". And, without any real mathematical argumentation, since he
adds "it is not mathematical problems that are important to discuss here". And he
 29

concludes: ("Instead I will limit myself to formulate considerations concerning this

particular problem")6. He concludes that his choice;

g11 = 1, g 22 = 1, g33 = 1, g 44 = − 1
(49)
gµ ν = 0 (µ ≠ ν)

presents itself, for him, as the only alternative representing "the only regular
solution of the basic equations of physics", and that it represents, according to
him, "a solution, and even the only regular solution of the basic equations of
physics".

Conclusion on this first part of Hilbert's 1916 article.

After the Second World War, at the turn of the seventies, a change of signature
was de facto ratified, through the scientific publications that followed, without an
article published in a physics journal justifying the reason. In the same way these
chimeras, like "the light cone", "outside of which" is a part of space qualified as
"elsewhere", populated with "spacelike curves", appeared, whereas this vision
comes from the projection of a reality associated with a hyperbolic geometry, in a
space of representation endowed with an elliptical geometry, an act of which we
have shown that it generated objects exempt from reality.
We had to go back to Hilbert's 1916 paper to trace the source of these drifts. In
fact, a "standard" vision of cosmology, populated by presentations that were
considered as acquired, not contestable, was built on the basis of later texts,
written in English, and therefore more easily assimilated in this language that
had become commonplace on a planetary scale after the war. The authors, like
medieval copyists, copied one another without any being able to return to the
fundamental texts if they did not master German.
To this we must add that if they had simply made the effort to go back to the
original version, by simply looking at equation (14) of this article they would
have been able to see immediately that their interpretation was in total
contradiction with the result of Schwarzschild.
It is significant that Hilbert's articles of 1915-1916 were only translated into
English in 2007 ([1], [2]), i.e. ninety years after they were published. Worse still,
these translations remain today, like many others, under copyright, as part of a
work gathering elements of this kind made available to scientists for a price
(October 2021) of 733 dollars for the printed version and 608 dollars for the

6In the English translation by Springer, 2007, we read : « is a mathematical

problem not to be discussed here. Instead I confine myself to presenting thoughts
concerning this problem in particular ». Le texte allemand est : «sind ist
mathematisch hier nicht allgemein zu erörternde aufgade. Ich beschränke mich
vielmehr darauf, einige besondere diese Aufgabe betreffende überlegungen
anzustellen ».
 30

digital version! We had to pay 58 dollars to buy the two pdf's corresponding to
the two articles that serve as a basis for this paper, while they are working tools
for researchers. If readers want to consult these translations, they will have to
pay the same amount. Today, there are no scientists, presenting themselves as
experts in cosmology, who have read the fundamental texts, or even know of the
existence of a mass of capital texts, some of which have not yet been translated
from German.
In what follows we will highlight a glaring and indisputable error of David
Hilbert, the immense impact of which will be measured in all the development of
cosmology that followed.

This error was first identified by the Canadian L.S. Abrams [7] in 1989, after
examining the original text, published in German by Karl Schwarzschild in
January 1916 [6], and comparing it with Hilbert's article of December 1916, also
in German.
In 1999 a similar approach was taken by A.Loinger [8], always starting from
German texts, which he reads fluently.

En 1999 l'italien S.Antoci, et l'allemand D.E Liebscher reprennent cette

question [9] et installent sur arXiv les traductions faites par Liebscher, enfin
disponibles en langue anglaise, 83 ans après leur publication en allemand, alors
que ces textes sont considérés comme la base même de la théorie des trous noirs.
En 2001 ils montrent que l'erreur relève d'une mauvaise compréhension de la
topologie de la solution trouvée par Schwarzschild [10] .
En 2003 S.Antoci réitère en publiant un article très documenté : David Hilbert
and the origin of the Schwarzschild Solution [11].

Plus récemment, en 2021, le Russe Anatoli Vankov souligne ce point en concluant

"Strictly talking the Black hole does not come from the general relativity theory"
[12].

Hilbert's error.
Let's go back to the chronology. On November 20, 1916 Hilbert published his
first paper entitled "Foundations of Physics" [1]. Five days later Einstein
published his own version of the field equation [4] as well as the first solution of
the linearized version of the equation, providing the first explanation of the
advance of Mercury's perihelion [5]. On December 22, 1915 the mathematician
Karl Schwarzschild, who avidly follows everything published in the section of the
Prussian Academy's annals devoted to mathematics, writes to Einstein
announcing that he has just constructed the non-linear solution of his equation,
which confirms his calculation. He announces to him that he will publish an
article in the same review. The text of this letter is available at the reference [3].:
 31

Fig.14 : Schwarzschild's letter to Einstein of December 22, 1916

Translation of the underlined passage:

R , θ , ϕ are not the coordinates "in use 7 », but it turns out that this
is the best way to express the metric.
We see that the genuine radial variable r is clearly visible. Schwarzschild
introduces an "intermediate variable" (Hilfsgröβe) R , according to the relation:

( )
1/3
R = r 3 + α3 ,

that Hilbert will confuse with a radial variable.

In his defense, we can say that in 1916 the mathematical tool "differential
geometry" was not completely mastered. In order to point out the error with
precision, we will resume Hilbert's calculation, point by point.
He begins by listing the hypotheses which are at the basis of this solution,
which he chooses to describe using what is for him a touchstone, Gaussian
coordinates, with:

(50) g14 = 0, g 24 = 0, g34 = 0,

The metric potentials are independent of time x4 . To this he adds that they
present a central symmetry (zentrisch symmeytisch), with respect to the origin of

7 Erlauten : authorized, permitted, standard, in use.

 32

the coordinates. An origin that he assimilated to the value R = 0 and not r = 0. In a

more modern way one would speak of an invariance under the action of the
group SO(3) and even O(3). Indeed, a geometric object which has this property
does not automatically have a "center". A torus has an "axial" symmetry, but this
axis only appears when we plunge it into ! 3 . Formally, this axis does not exist.
Similarly, returning to the metric of the "3D diabolo", this object does not have a
"central symmetry", because this "center" only appears when we project it into a
representation space which is the three-dimensional Euclidean space. It is
invariant by action of the group O(3).
Hilbert then writes:
(51)
In agreement with Schwarzschild, if we pose:

w1 = r cosϑ
w2 = r sinϑ cosϕ
w3 = r sinϑ sinϕ
w4 = l

We find a character l which translates the vision that Hilbert has of the universe.
These are four coordinates which are, let us refer to the beginning of his article,
universal parameters (weltparameter) ( w1 , w2 , w3 , w4 ) . What is extraordinary is
that Hilbert will conduct his calculation in this coordinate system. The time t will
appear only at the very end "when all the cosmic mechanics will start working".
In the Lagrange equations, which will follow, he does not manipulate the
dt dp
derivative mais la dérivée .
dp dp

This is a way for him to affirm that this metric "exists", that this geometry "pre-
exists" before God decides that l = it, before starting the time race. For Hilbert the
solar system exists, as it is, since the creation of the universe by God, since time
zero.

Another remark, in this following you will find g and not − g . So the
determinant of the metric is positive. We have to remember that for Hilbert the
curvature phenomena are exceptional accidents, almost imperceptible folds in a
practically Euclidean universe. The universe associated to these coordinates
( w1 , w2 , w3 , w4 ) is therefore Euclidean. Does it have a length? No, not yet. It is
only when this length (these lengths in the plural, if we stick to Hilbert's text,
which defines two of them) that the space becomes pseudo-Euclidean and that
the determinant becomes negative. Before God gives it a physical character, it is
an object of pure mathematics. We could call it "metaphysical"..
 33

-> When Einstein produced his field equation, he created a tool with which
to interpret physical phenomena, accessible to astronomy. He is thus
already in a four-dimensional world ( x1 , x2 , x3 , t ) , concret. The
determinant of its metric is therefore negative and it must handle a −g .

-> When Hilbert produces his own version of the field equation he is in
another universe, which he wants to be more abstract, more fundamental,
the universe of ( w1 , w2 , w3 , w4 ) . You now have an explanation for the
presence of the g in its field equation, constructed and written in the
system ( w1 , w2 , w3 , w4 ) .

It is extremely regrettable that I cannot provide a link to the English translation

of Hilbert's 1915 article [1], which is still under this scandalous copyright, even
though it is a key element of the world's scientific culture. So it will cost you 29
dollars if you want to check that the coordinates ( x1 , x2 , x3 , x4 ) are totally absent
∂
from this article where all derivatives are in . This is true for all terms, the
∂ws
Christoffels coefficients, the terms of the Ricci tensor.

So the Hilbert field equation refers to a space ( w1 , w2 , w3 , w4 )

End of this digression. In his 1916 paper Hilbert puts the bilinear form in the
form :

(52) F(r) dr 2 + G(r)( dϑ 2 + sin 2 ϑ dϕ 2 ) + H (r) dl 2

A further step has been taken. Hilbert is then in a reference frame:

(53) { w = x , w = x , w = x , w =l }
1 1 2 2 3 3 4 1

The passage in polar coordinates implies that its variable r is defined, as with
Schwarzschild [10]8; by :

(54) r= x12 + x2 2 + x32 ≥ 0

Then, thanks to the change of variable, implicit, it is passed in polar coordinates:

(54) ϑ = arcsin x3 ϕ = arccos x1

We are thus in the coordinates (51). From (52) it is clear that Hilbert expresses
his bilinear form in the system:

(55) ( r , ϑ , ϕ, l )

8 What you can check in the English translation, not covered by a copyright
 34

Other form of the system ( w1 , w2 , w3 , w4 ) .

There is no question of time.

He then poses :

(56) r * = G(r)

It is then that he will commit a major error, reported since 1989 in

([7], [8], [9], [10], [11], [12]) .
He writes:

Fig.15 : Hilbert's error

Translation :

- We are therefore, in the same way, entitled to interpret ( r * , ϑ , ϕ, l )

as polar spatial coordinates. If we introduce r instead of r* in (our
expression of the bilinear form) and omit the * sign again, we obtain
the expression:

(57) M (r) dr 2 + r 2 ( dϑ 2 + sin 2 ϑ dϕ 2 ) + W (r) dl 2

This bilinear form is a solution of the Einstein field equation without a second
member which then reduces to canceling the components of the Ricci tensor,
which Hilbert denotes by K µ ν . These are calculated on the basis of Christoffels
symbols. Note that Hilbert, as in his 1915 paper, does all his calculations with the
 35

variables ( r , ϑ , ϕ, l ) i.e. "universal"- coordinates » ( w1 , w2 , w3 , w4 ) . His

calculation of the geodesics is then based on the variation of the action,
constructed with these same variables :

⎛ ⎛ dr ⎞ 2 2 2
⎛ dl ⎞ ⎞
2
2 ⎛ dϑ ⎞ 2 ⎛ dϕ ⎞
(58) δ ∫⎜ M ⎜ ⎟ +r ⎜ + r sin ϑ ⎜
2
+ W ⎜ ⎟ ⎟ dp = 0
⎜⎝ ⎝ dp ⎠ ⎝ dp ⎟⎠ ⎝ dp ⎟⎠ ⎝ dp ⎠ ⎟⎠

Hs corresponding Lagrange equations are:

(59)

The type ' in these equations, and in what follows, refers to a derivation with
respect to r. The differential equations of the geodesic curves are:

d 2 ws ⎧ µ ν ⎫ dwµ dwν
(60) 2
+∑ ⎨ ⎬ =0
dp µν ⎩ s ⎭ dp dp

We notice that we are always in these "universal" coordinates ( w1 , w2 , w3 , w4 ) .

Hilbert then computes non-zero Christofells symbols:

(61)
 36

This allows him to calculate the components of the Ricci tensor 9

(62)

The calculation of the Ricci scalar follows:

9In the English translation, under copyright, a sign error in the equation
giving K22 where the translator has put a plus sign instead of a minus sign in
the second term of the right member.
 37

(63)

Hilbert places himself in a coordinate system where the determinant is positive,

which allows him to write:

(64) g = M W r 2sinϑ

Then :
(65)

He poses

r r-m
(64) M= W = w2
r-m r

-> The letter w does not designate the modulus of the vector ( w1 , w2 , w3 , w4 )

It is an unknown function. By making the change (64) Hilbert will now have
two unknown functions to determine: m and w . Why such a change? It is inspired
by the equation in Schwarzschild's paper and these functions m and w will turn
out to be simple constants.

But, in passing, we discover the origin of lette m used to describe what has the
dimension of a length!

It comes from:
(65)

Hilbert constructed his action on the basis of a function H = K + L , where K is

the Ricci scalar. But in a portion of the universe which is empty, L = 0 . So the
variation reduces to:

(66) δ ∫∫ ∫∫ K g dr dϑ dϕ dl = 0
 38

Note that we are still in the coordinate system ( r , ϑ , ϕ, l ) . These geodesics

"exist", but "as God has not yet created time", the planets cannot launch
themselves on these geodesic-orbits. This equation is equivalent to:

(67) δ ∫ wm' dr = 0

And the Lagrange equations then give:

m' = 0
(68)
w' = 0

The solution constructed by David Hilbert is then written:

r r−α 2
(69) G ( dr , dϑ , dϕ , dl ) = dr 2 + r 2sin 2 ϑ dϕ 2 + dl
r−α r

And, posing l = i t (i.e. w4 = i x4 , according to Hilbert's notation).

r r−α 2
(70) G ( dr , dϑ , dϕ , dt ) = dr 2 + r 2sin 2 ϑ dϕ 2 − dt
r−α r

In the original text we find a typographical error of Hilbert who leaves his
"universal variable" l in the first member 10.

Fig.16 : A typographical error by Hilbert.

At this point, Hilbert is convinced that he has found Schwarzchild's result. He
writes:

10 Error also in the translation published by Springer (still under copyright): the
letter l must be replaced by the letter t in both members. Considering the number
of obvious errors in this translation, available since 2007, that is to say, at the
time of writing, since 14 years, it is doubtful that this translation has been read by
people mastering general relativity.
 39

- Für l = it die gesuchte Maβbestimmung in der von Schwarzschild zuerst

gefunden Gestalt
Translation :
- For l = it we find the metric first constructed by Schwarzchild.
It is this error that was very quickly propagated through successive erroneous
interpretations.

The true Schwarzschild metric.

Had Schwarzschild survived that spring of 1916, when he died of an infection
contracted on the Russian front, he would have immediately brought these
commentators of his work back to the original form of it, where the nature of R is
well specified and corresponds only to an intermediate quantity (Hilfsgröβe) and
in no case to the radial distance r.

Fig.17 : Schwarzschild's true metric, 1916

What is extraordinary is that this confusion between R and r, which is the basis
of Hilbert's error, has been a standard for more than a century, even though it
was obvious in Schwarzschild's result in equation (14), even to a non-German
speaker. The only possible explanation was that this misinterpretation spread,
like the misinterpretation of a founding text by people behaving like medieval
copyists.
In fact, the generations of theorists that followed were satisfied by the fact that
this solution fulfilled for them a consideration that could be considered as
fundamental: to identify with the Lorentz metric at infinity. In his article of 2003
[11], the Italian mathematician Salvatore Antoci analyzes with the greatest
precision the mechanism of construction, by Hilbert, of this error, putting in
perspective his calculation and that of Schwarzschild. In another paper [10] ,
Antoci and the German D.E. Liebscher show that Hilbert's error translates an
erroneous interpretation of the topology of the object.
 40

To this we must add that for 43 years the only application that came to mind
concerned the linearized form of this metric. And this is the reason why
Schwarzschild himself does not provide the expression of its nonlinear solution
in his coordinates. In his paper, Schwarzschild takes the case of the Sun, where
the large α is then 3 km , noting:

Fig.18 : Schwarzschild justifies his limitation to the linearized solution.

Translation:
- Nevertheless, Mr. Einstein's approach to the calculation of the geodesics is
compatible with the exact solution, if we express it using r instead of:
1/3
⎛ α3 ⎞
( )
1/3
R= r +α 3 3
= r ⎜ 1+ 3 ⎟
⎝ r ⎠

since α/r is close to twice the square of the planetary speed (taking the
speed of light as the unit). For Mercury, the order of magnitude is 10-12.
Thus R is practically equal to r and Mr. Einstein's approach is sufficient,
beyond the needs of current practice.

If Schwarzschild had finalized his work, which he did not find necessary, which
he would have done by expressing the metric in the coordinates ( r , ϑ , ϕ, t ) this
would have led him to write the true solution, which is obtained immediately by
performing the change of coordinate, explicitly mentioned by Hilbert. An
operation which would have made appear this true metric solution of the
Einstein equation without second member.
 41

(71)

( r 3 + Rs3 )1/3 − Rs 2 2 r4
ds 2 = c dt − dr 2 − ( r 3 + Rs3 )2/3 ( dθ2 + sin 2θ dϕ 2 )
( r 3 + Rs3 )1/3 3 3
⎡ 3 3 1/3
⎤
( r + Rs ) ⎣( r + Rs ) − Rs ⎦
r ≥0

In a footnote on page 70 of his manuscript, Hilbert signs his obvious

misunderstanding of the relation (number 14 without Schwarzschild's original
paper; which accompanies his expression of its metric solution, according to the
intermediate quantity R :

( )
1/3
(72) R = r 3 + α3

Fig.18 : Hilbert’s footnote

Translation :
- In my opinion I will not recommend, as Schwarzschild does, this
transformation bringing the point r = α to the origin, especially
since there are simpler ways to achieve this.

Thus, while the expression according to the true radial coordinate r is

consistent with the approach followed by Schwarzschild, the fact of presenting
the result according to this intermediate quantity R is only an artifice used by
Schwarzschild to stick more simply with the linearized solution of Einstein,
Hilbert reverses the reasoning by considering R as the radial variable and r as an
artifice of calculation to get rid of the singularity in what he believes to be the
origin of the coordinates, in R = 0 , whereas, according to (72) this point
corresponds to the pure imaginary value:
(73)

( )
1/3
r=i R3 − α 3
 42

In doing so, Hilbert does not realize that he has left the domain of definition of the
manifoold M4.
An error that will lead to floods of ink and the production of theorems
(Penrose, Hawking) referring to a "central singularity", existing only in the
imagination of scientists who have studied this question, thus taking for a reality
what belongs, mathematically speaking, to an imaginary space domain.
Coming back to this expression (71), by performing a series development we
obtain the values of the metric potentials when r tends to zero:

3r
(74) gt t → 0 gr r ! →0 gθθ → Rs 2 gϕϕ → Rs 2sin 2θ
Rs

At r = 0 the Kretschman scalar is nonzero, so the sphere is not a singular locus.

On the other hand the determinant is zero, which indicates that the hypersurface
is locally inorientable. It is doubly so, since the two potentials and are zero. If the
sphere is a gorge sphere, it reflects a double inversion of space and time, a PT-
symmetry, as established in . [13]]. If we consider this nullity of the determinant
as a singular region" we can then conclude that the hypersurface is not a 4-
manifold but a 4-orbifold.

Recall, as Hilbert notes, that these coordinates are Gaussian. We can therefore
consider a layering of the hypersurface where the variable with t as parameter.
This corresponds for these three-dimensional hypersurfaces to:

r4
(75) dσ 2 = dr 2 + ( r 3 + Rs3 )2/3 ( dθ2 + sin 2θ dϕ 2 )
( r + Rs ) ⎡⎣( r + Rs ) − Rs ⎤⎦
3 3 3 3 1/3

These are themselves subject to a new layering at r = Cst which gives a family of
nested spheres like Russian dolls having the minimal area, corresponding to r = 0:

4 π Rs
2
(75)

The 4D hypersurface is therefore non-contractile. It is then :

- Either a bordered manifold.

- Either a geometrical object translating a space-time bridge between two
Minkowski spaces, realized through a sphere of throat area 4 π Rs2 .

We can produce a finer description of the object by introducing a variable

through the change of variable [13]. :

(77) r = Rs ( 1 + L n chρ )

The metric then becomes:

 43

(78)

Logch ρ 2 2 1 + Logch ρ 2
th ρ d ρ 2 − Rs2 ( 1 + Logch ρ ) ( dθ )
2
ds 2 = c dt − Rs2 2
+ sin 2 θ dϕ 2
1 + Logch ρ Logch ρ

The determinant is always zero because of the term gt t but now gρρ → 2 .

This explains why theorists have massively embarked on this erroneous

interpretation of Schwarzschild's solution, which is repeated on several points.
-> The fact that the non-linear aspects of the solution were taken into
account only in 1939, after the publication of a key article [14] by R.
Oppenheimer, without references, which we will mention in the following.

-> The fact, highly probable, that no cosmologist, starting from this
"founding article" had the curiosity to take a look at the original article, in
German, in which case the equation (14) of this paper should have attracted
their attention.

-> The fact that this curiosity came up against the fact that these German
texts were only available, until the advent of the internet and pdf files, in
books that were generally expensive.
-> The fact that Schwarzschild's article was only translated into English in
1999 [6], i.e. 86 years after its publication in its original form.
-> The fact that the translations of the documents that are essential to clear
up this matter were only translated into English in 2007 ([1], [2]), that is 91
years after their publication, and are still covered by a scandalous copyright.

-> In 1960 the publication by M.D.Kruskal [15] of the construction of an

analytical extension allowing to "penetrate inside the Schwarzschild
sphere" gives the illusion of a progress in the understanding of this
geometry, with that it does not change anything to the case. This extension,
which refers to an imaginary r-value, is outside the definition space of the
geometric object. In the same way one could build an analytical extension
allowing to study a torus inside its throat circle.

-> To this we must add, for more than half a century, the illusory conviction
of a progress in the understanding of this solution, whereas the generalized
contagion of this flagrant error has made thousands of publications fall into
what is nothing but science fiction.

The birth act of the black hole model.

In 1939 the hypothesis of the existence of neutron stars was already circulating,
although their existence was only confirmed in 1967 by the discovery of pulsars.
In their article [14] J.R. Oppenheimer and H.Snyder explicitly mention them. In
 44

fact, it all started with an article published a few months earlier by R.C. Tolman,
then working at Caltech [16].

Fig.19 : Richard Tolman and Albert Einstein

We find this expression of the metric:

Fig.20 : Tolmans’ line element.

This description still places us in Gaussian coordinates. We can therefore leaf

through the space-time by using t as a parameter, translating a simple temporal
 45

translation and by considering the three-dimensional hypersurfaces described by

the metric:

(79) dσ 2 = e λ(r ) dr 2 + r 2 dθ2 + r 2sin 2θ dϕ 2

Hypersurfaces that can be foded through a family of area spheres 4 π r 2 . Can

this area be brought to zero? If so, this will mean that the geometric object
associated with this portion of space-time is contractible.

Moreover, by introducing the two functions e ν(r ) and e λ(r ) strictly positive, if we
are in a mode governed by reals, Tolman rejects a possible modification of the
hyperbolic signature which, for him, is ( - - - + ); or ( + - - -) according to the order
of the terms.

Fig.21 : Tolman's expression after Oppenheimer and Snyder.

This work is in fact a resumption of the second article [17] published by

Karl Schwarzschild, just before his death, where he entirely constructs the
geometry inside a sphere filled with an incompressible matter, of constant
density. R.Oppenheimer and H.Snyder quote him in the article they published
in 1939 [14] and start again from the same form of the metric, this time to
take up the question of geometry outside the mass. The question of the end of
life of stars having exhausted their fusion fuel is at the center of the article.
The accentuation of the gravitational redshift is evoked, as the contraction of
the star increases. The free fall time of a test particle is calculated in two ways.
The authors show that if we rely on the proper time, this time is finite, and
very short. On the other hand, if we measure this time by using the coordinate
t , which is supposed to refer to the proper time of an observer located at a
great distance from the object, this time becomes infinite.

This remark signs the birth of the Black Hole model, according to the
following reasoning:
 46

-> A star, no longer able to counterbalance the force of gravity with the
help of pressure, undergoes a free fall towards its center, which
nothing can counteract.

-> Without undertaking a description of this phenomenon, we rely on

the data emanating from the metric describing the exterior of this
object, which we will call "Schwarzschild's exterior metric".
-> By calculating the free fall times of test-mass, if we find that it
reaches the Schwarzschild sphere in a finite time, this time becomes
infinite by an observer located at a great distance. For the latter, this
implosion phenomenon seems to be like a frozen.

-> At the same time, the radiation emitted by the matter undergoes a
gravitational redshift effect which becomes infinite when this signal is
emitted from a point located on the Schwarzschild sphere of radius Rs.
Thus, a fortiori, no radiation can cross this sphere which will be
qualified as cosmological horizon.
An outside, distant observer will then perceive this object as a perfectly black
disk, which will be called a black hole.
This reasoning allows us to free ourselves from the description of the collapse
phenomenon by starting from the reasoning:
-> I do not feel obliged to describe a phenomenon which for me,
an outside observer, lasts an infinite time.

This also allows us to reduce the description of the geometry of the object to
the only geometry referring to the outside of the horizon sphere, thus to a
solution of Einstein's equation that refers to a portion of the empty universe.
Assuming that one can start from the solution of figure 21 to calculate the free fall
time of a witness particle up to the point r = 0 , supposed to be the "center" of the
object, one obtains a finite and brief value. We deduce, although we cannot carry
out any observation on what happens and has happened inside the horizon
sphere, that all the matter is concentrated in a central singularity.

This reasoning is based on the assumption that the considered expression of

the solution has a physical meaning. However, as mentioned above, this is not the
result found by Schwarschild in 1916 but the result of the error committed by
( )
1/3
Hilbert, by confusing the intermediate quantity R = r 3 + α 3 with the radial
distance r. Considering to exploit a calculation referring to a value r < ro
( Schwarzschild radius) one is simply outside the four-dimensional hypersurface,
which can be seen immediately from the fact that the exponential functions
become negative. Now :

(80) eλ = − n → λ = Ln n + iπ
 47

A complex function appears, itself a function of a complex value of the variables.

Thus the supposed "interior" of such an object exists only in the imagination of
theorists, in the strict sense, since they decide to consider as real what is
imaginary.

The emergence of surrealism in physics.

The years have passed. No theorist cares to return to the founding texts, nor to
consider another model. Here are the arguments that appear in all the books and
manuals intended for the training of students. As an example, we reproduce
elements of section 6.8 of chapter 6 of the book of reference [18]. The choice of
the form of the metric introduced by Tolman [16] is then simply presented as
"reasonable".

When r becomes less than 2I (the Schwarzschild radius), the signs of the
components of the metric (potential metric referring to time) g11 ( metric
potential referring to the assumed radial coordinate) change, g11 becomes
positive ans g oo negative. This forces us to reconsider the physical meaning (...)
given to the variables t and r as a system of marking time and radius, inside the
Schwarzschild sphere. In fact a line of universe which is described according to a
t , that is to say with ( r , θ , ϕ ) constants, corresponds to ds 2 < 0 . It’s a spacelike
curve, while a line of universe for which ds 2 > 0 is a timelike curve.

And there you come across the consequence of another of Hilbert's errors, that
of endowing space-time with two systems of measurement, the second referring
to portions of curves that he calls "segments" and for which the sign of the
bilinear form is inverted.

Fig.22 : The second "length" measured on Hilbert « segments ».

 48

Translation :
- A portion of a curve where (the form G is positive) will be called a
segment, while (the expression giving the scalar λ ) is the length of
this segment.
This vision of things is in total contradiction with the one of Einstein, K.
Schwarzschild, J. Droste, H.Weyl, and all the scientist-mathematicians who at that
time contributed to the construction of the general relativity. In the article of
Schwarzschild, for example, we read:

Fig.23 : How Schwarzschild defines length, essentially positive.

Translation :
- Consider a point that moves according to (the expressions in the
figure), where the variables are functions of the x variables
(coordinates of points of the space-time hypersurface) and where the
values of x must be considered as constant at the beginning and at
the end of the path followed for the integration. Clearly, the point will
have to move according to a geodesic of the variety (manifold)
characterized by the element ds.

How did Hilbert come to endow space-time with two lengths? Perhaps he had in
mind to create a metaphysics, with respect to the events occurring inside (...) the
Schwarzschild sphere.

Let's go back to the text of the reference [18].

- Let us take again the text of the reference It would thus appear
natural (...) to treat r as a time coordinate and t as a radial
 49

coordinate (...). We interpret ds/c as the proper time along the

lines of universe traveled by a particle. Then, as we have shown in
section 4.2, this definition only has a physical meaning if .
Similarly, a massive particle cannot maintain itself on a
trajectory with constant r inside the Schwarzschild sphere, which
would imply that ds 2 < 0 along this line of universe.

Let us quote the "standard" derivation of the so-called Schwarzschild solution,

e.g., by quoting the corresponding pages of chapter 6 of the 1975 book of
reference [18]. Page 186 equation (6.4) represents the most general form of the
metric, in the absence of cross terms. The minus sign shows that the authors
intend, with A , B , C ; D positive, to introduce the signature of the metric right
away:

Fig.24 : Excerpt from page 186 of the reference [18]

But in the last lines we read "nevertheless, it is possible to obtain an additional

simplification by a judicious choice of radial coordinate"
 50

Fig. 25 : Excerpt from page 187 of the referenc [18].

And there we see, reproduced identically, the reasoning held by Hilbert in his
1916 paper, which shows that he is at the origin of this distortion of the true
Schwarzschild solution. In equation (6.9) the authors take up the introduction, in
1939, of two exponentials by Tolman [16] and Oppenheimer [14]. They even
specify that these functions are "intrinsically positive" and that "these
coordinates have a clear physical meaning", whereas, precisely, attributing a
physical meaning to coordinates is the first source of error.

Six pages later, on page 193, we find the result of their calculation:
 51

Fig. 25 : Extrait de la page 193 de la référence [18].

If these exponential functions are "intrinsically positive", then the variable r

cannot be less than 2m, otherwise the quantities and would correspond to:

2m 2m
(81) eλ = − 1− → λ = Ln 1 − + iπ
r r

The artifact of the Kruskal coordinates.

In spite of this obvious contradiction D.Kruskal has constructed an analytical

extension, so as to be able to build a description of this "interior of the
Schwarzschild sphere ( 0 < r < 2m ). The reader will find the "standard"
construction of these new coordinates u and v , as well as the resulting metric, in
pages 226 to 230 of the reference [18]. We will only reproduce the equations
themselves. First, we have the two equations identified by (6.91) in the book:

r
(82) ξ = r + 2 m Ln −1
2m

1 − 2m / r
(83) F (ξ) =
f2

From these relations it is established by introducing the intermediate quantity

η , equation (6.200) of reference [18] shows that:

(84) F (ξ) = η2 e 2 ηξ

We then read, in the equations (6.201) :

 52

(85)

2mη
⎛ r ⎞
u=⎜ −1⎟ e η r ch ηx°
⎝ 2m ⎠

2mη
⎛ r ⎞
v=⎜ −1⎟ e η r sh ηx°
⎝ 2m ⎠

1− 4mη
2m ⎛ r ⎞
f = 2 ⎜
2
−1⎟ e −2η r
η r ⎝ 2m ⎠

Then:

Fig. 26 : From page 229 of the reference [18].

The scalar m being real, η is therefore also real.

This leads to the final expression of the new variables, in two configurations.

(86) r > 2m :

r x°
u= − 1 e r / 4m ch
2m 4m

r x°
v= − 1 e r / 4m sh
2m 4m

32 m3 r / 2m
f2= e
r

⎛ r ⎞ v x°
(87) u2 − v 2 = ⎜ − 1 ⎟ e r/ 2 m = th
⎝ 2m ⎠ u 4m
 53

And

(88) r < 2m :
r x°
u= 1− e r / 4m sh
2m 4m

r x°
v= 1− e r / 4m ch
2m 4m

32 m3 r / 2m
f2= e
r

With :
⎛ r ⎞ r/ 2 m u x°
(89) v 2 − u2 = ⎜ 1 − e = th
⎝ 2 m ⎟⎠ v 4m

The metric then takes the form, equation (6.187) of the reference [18]:

(90) ds 2 = f 2 (u, v ) ( dv 2 − du 2 ) − r 2 ( dθ2 + sin 2θ dϕ 2 )

Although this metric is not identified with the Lorentz metric at infinity, when
we form ds = 0 we have , see equation (6.188) of reference [18] :

2
⎛ du ⎞
(91) ⎜⎝ dv ⎟⎠ = 1

Here is the famous Kruskal diagram:

 54

Fig.27 : Kruskal diagram.

If one follows this diagram, where all quantities become real, with a real ds one
thus manages to penetrate "inside the Schwarzschild sphere". The points located
at constant r are on hyperbolas. The one on the right refers to the value r = 4m.
The path A is that of a particle with mass, which plunges towards the
Schwarzschild sphere. The half line x° = 2m evokes the way time evolves. In the
chosen coordinate system this mass reaches this sphere in an infinite time,
shown by the half line v = u . Then, assuming that this point-mass can penetrate
inside this sphere, it continues its way and reaches the point associated with a
zero value of r which is another hyperbola accompanied by hatching, which is
supposed to represent "the singularity". A similar reasoning is associated to the
trajectory B.

By a real magic wand, Kruskal seems to have transformed an unreal portion of

the variety into something real, described by equations (85) to (89). But we have
to remember that if with real quantities, like space-time coordinates, we can
obtain a complex quantity ds, the opposite is also possible.

So we have to go back to Kruskal's approach. Let us explain the relation (84)

1
(92) F (ξ) = η2 e 2 ηξ = 2
e ξ/ 2 m
16 m

Within the Schwarzschild sphere (83) indicates that:

1 − 2m / r
(93) F (ξ) = <0
f2

Combined with the previous equation this gives us:

 55

(94) e ξ/ 2 m = 16 m2 F (ξ) < 0

Now the exponential can only be negative if the exponent is complex:

(95) ξ = 2 m L n ⎡⎣16 m2 F (χ) ⎤⎦ − i π

Note all the calculation has been based on the hypothesis (82), which leads to:

r
(96) ξ = r + 2 m Ln − 1 = 2 m L n ⎡⎣16 m2 F (χ) ⎤⎦ − i π
2m

There is therefore a contradiction. This relation becomes incoherent. In the two

members of an equation, one cannot be real and the other complex. This
analytical extension makes sense if we are in the mode of complexes, but it does
not make sense in the mode of physics, which is in the world of reals.

The interpretation of Hermann Weyl ( 1917)

In 1915 Hermann Weyl was thirty years old. After having taught mathematics
at the University of Göttingen, where he was fascinated by the revolutionary
ideas introduced by Riemann and particularly by hyperbolic varieties, he found a
position in Zurich, at the Federal Polytechnic, where he was offered a chair. He
then met Einstein and quickly assimilated the basic concepts of relativity, first
special and then general. Discovering the exact nonlinear solution found by Karl
Schwarzschild, he published his interpretation in 1917 [19].

Fig.28 : Hermann Weyl in 1915

 56

To access the German version of this article, published more than a century
ago but copyrighted by Springer in a 2012 republication which also includes the
English translation, it will cost you $49, regardless of the version. In this paper,
unlike Hilbert's, the letter R is used to designate the Riemann scalar:

Fig.29 : Derivation of the field equation by Weyl.

The first equation shows that Weyl immediately integrated the technique of
derivation of the field equation by the variational method, with the introduction
of the Ricci tensor and the scalar R derived from it. At the bottom, we see the field
equation to which Einstein gave his name, in the form (equivalent to that of his
publication of November 25, 1916 [4]) that students know today. The first thing
that Weyl does is to remind us of the inequality which gives the solution a
physical character:
 57

Fig.30 : Length measurement, after H.Weyl [21].

Like Schwarzschild, he is perfectly clear in his choice of Gaussian coordinates

( x1 , x2 , x3 , x4 ) . The space-time is therefore foldable by means of a sequence of
three-dimensional hypersurfaces, invariant with respect to time x4. The problem
is to construct this stationary three dimensional hypersurface, described by the
coordinates ( x1 , x2 , x3 ) and defined by its length element dσ according to :
(97) ds 2 = f dx4 2 − dσ 2

If we compare the choices made by Weyl with Schwarzschild's approach, the

latter carries out his calculation with a coordinate r which is in fact his
intermediate quantity R. He calculates the function f as well as the expressions of
the three other metric potentials.

Fig.31 : Result of the calculation of the potential g 44 = f [21].

Like Schwarzschild, he makes appear what he calls the gravitational radius

associated to a mass m. His metric can be summarized as:

⎛ 2a ⎞ dr 2 ⎛ 2a ⎞
(98) ds 2 = ⎜ 1 − ⎟ dt 2 − − r 2 ( dϑ 2 + sin 2 ϑ dϕ 2 ) = ⎜ 1 − ⎟ dt 2 − dσ 2
⎝ r⎠ α ⎝ r⎠
1−
r
 58

With the 3D hypersurface 11 defined by the metric:

dr 2
(99) dσ =
2
+ r 2 ( dϑ 2 + sin 2 ϑ dϕ 2 )
2a
1−
r
But Weyl is not naive enough to consider that this solution could make physical
sense for r < 2a, where the length element would cease to be real. This variable r
(which is not the same as the one in the Schwarzschild paper) corresponds to a
ray vector with coordinates ( x1 , x2 , x3 ) . It is a 3-dimensional hypersurface which
we know "exists" only for r > 2a. Weyl will therefore push further, to pierce the
secret of its geometry, and more precisely of its topology. On these three space
coordinates he can always delete one of them. So he writes:

Fig. 32 : The analysis of the topology of the 3D hypersurface by Weyl

11 We recognize the metric of our "3D diabolo" from the beginning of the article.
 59

Then he weites:
- In order to determine the geometry which is characterized by the
form of the metric giving dσ 2 , we will project in a corresponding
plane à x3 = 0 . If we introduce the polar coordinates:

x1 = r cosϑ x2 = r sinϑ

dσ 2 = h dr 2 + r 2 dϑ 2

Ce qui lui donne :

⎛ 2a ⎞
dσ 2 = ⎜ 1 − ⎟ dr 2 + r 2 dϑ 2
⎝ r ⎠

In the following, it does exactly what we did for the 3D diabolo, i.e. it sets the
angle ϑ to determine the equation of the meridian, which gives it the differential
equation:

⎛ 2a ⎞
(100) dr 2 + dz 2 = ⎜ 1 − ⎟ dr 2
⎝ r ⎠

Whose solution is:

z2 z2
(101) z= 8a ( r − 2 a ) or r = 2 a + = Rs +
8a 4 Rs

We find the equation(19) of the lying parabola. The hypersurface is thus a "3D
diabolo" which is the projection of the 3D hypersurface in a Euclidean space
( x1 , x2 , x3 ) . And Weyl adds :

- Die projektion bedekt das äubere doppelt : This projection covers

twice (doppelt ) the portion of space r > 2 a , das innere überhaupt
nicht : but this (3D) structure definitively does not fit the portion of
space r < 2 a 12 . BeI natürlicher analytischer Fortsetzung wird also
der wirkliche Raum in dem zur Darstellung benuntzen Koordinaten
der xi das durch r ≥ 2a gekennziechnete Gebeit dopplet überdecken :
In a natural analytical extension, the real space in the coordinates
used for the representation of the coordinate xi of the point
correspond two points of the 3D hypersurface. On this sphere of
radius 2a, which makes the junction between these two coverings,
is located the mass.

Là on voit poindre une idée tout à fait extraordinaire, assimilant les masses à
des singularités topologiques. On se rappellera que la topologie, à cette époque,

12On dirait aujourd’hui : cette hypersurface 3D constitue le revêtement à deux

feuillets de la portion d’un espace Euclidien 3 extérieur à une sphère de rayon 2a.
 60

est en train de naître en tant que discipline mathématique. Les surfaces fermées
régulières 2D sont au nombre de quatre. On a la sphère, le tore, puis la bouteille
de Klein et la surface de Boy. Précisos que Félix Klein invente sa bouteille en
1882, tandis que Werner Boy, élève de Hilbert, créera sa propre surface en
1902. Comme Schwarzschild ce dernier s’engage à 35 ans, dès l’entrée en guerre
de l’Allemagne, en juillet 1914 et est tué en France, où il repose, en septembre
de la même année.
Weyl (who also joined the army, but was discharged for health reasons)
continued his analysis of the 3D hypersurface. He is thus the first to introduce the
isotropic form of the metric.

Fig. 33 : Isotropic form of the Schwarzschild metric.

His variable r is not the previous one. Its formula (12) corresponds to equation
(6.69) of reference [18]. The last expression represents the linear expansion
coefficient (lineare Vergröberungsverhältnis).
It is clear that Weyl has perfectly integrated the fact that the various
coordinates are only representations of the objects defined by their metrics and
that the only object endowed with an intrinsic reality (invariant by any change of
coordinates) is the element of length s . These successive choices allow us to
discover their topology. Thus, in 1917 is the first to discover that the geometry
discovered by Schwarzschild is non-contractile.
 61

The P-symmetry that goes with the passage of the throat sphere.
Weyl thus creates this concept of a representation space and a two-sheet
covering of a manifold. The homologous points of these two four-dimensional
sheets can thus be identified using the same coordinates xi. We can consider four
points in the vicinity of a point of coordinates xi which form a tetrahedron
consisting of four equilateral triangles having two common vertices. We can
define a positive orientation by defining a direction of travel of these triangles,
considered as positive. This one defines a normal vector. The following figure
shows what happens to this set of points when they cross the throat sphere. The
tetrahedron with black edges is supposed to belong to one of the two
tridimensional layers of the object. Let one of its faces ABC, the positive direction
of travel, arbitrary, being indicated by arrows. The tetrahedron with the shaded
edges A'B'C' belongs to the other layer. If we bring these two objects into
coincidence we can see that the two directions are the opposite of each other

Fig.34 : Reversal of the orientation of a tetrahedron after crossing the throat

sphere.

We deduce that any object crossing the throat sphere of the Schwarzschild
geometry undergoes a P-symmetry.

Free fall time or escape time in Schwarzschild geometry.

The black hole model is based on the complete decoupling of the proper time
of objects accompanying the implosion phenomenon that this geometry is
supposed to describe and the proper time of an observer located at infinity,
observing the phenomenon, which for him is supposed to last an infinite time. Let
us consider radial trajectories. We have (keeping the Schwarzschild notation) :
 62

⎛ Rs ⎞ dR 2
(102) ds = ⎜ 1 − ⎟ dx4 −
2 2
s = cτ x4 = ct
⎝ R⎠ R
1− s
R

The Lagrange equations give:

dτ 1 R
(103) =±
dR c Rs

dt 1 R
(104) =±
dR c R - Rs

The integration gives the diagram:

Fig. 35 : The free fall time.

The witness particle reaches the throat sphere in a finite time, in terms of its own
time, but this path corresponds to an infinite time for a distant observer.

The Kerr metric.

In 1963 Roy Kerr constructed the solution of the Einstein equation without a
second member [20], describing a portion of empty space, invariant by time
translation and by action of the group O(2).
 63

(105)

⎛ 2 mρ ⎞ 2 2 ρ 2 + a 2 cos 2 θ
ds = ⎜ 1 − 2
2

⎝ 2 2 ⎟
ρ + a cos θ ⎠
c dt − 2
ρ + a − 2mρ
2
dρ2 − (ρ 2
)
+ a 2 cos 2 θ dθ 2

⎡ 2 m ρ a sin θ ⎤ 4 m ρ asin θ
( )
2 4 2
− ⎢ ρ 2 + a sin 2 θ + 2 ⎥ dϕ 2
− c dt dϕ
⎣ ρ + a 2 cos 2 θ ⎦ ρ 2 + a 2 cos 2 θ

The parameter a figures the importance of the rotation. If it is zero we find the
Schwarzschild metric. In the plane this metric becomes:
(106)

⎛ 2m ⎞ 2 2 ρ2
ds = ⎜ 1 −
2
⎟ c dt − 2 dρ2
⎝ ρ ⎠ ρ + a − 2mρ
2

⎡ ⎤ 2 4m a
( )
2
2m a
− ⎢ ρ2 + a + ⎥ dϕ − c dt dϕ
⎣ ρ ⎦ ρ

Consider two rays of light emitted tangentially to a trajectory at ρ constant :

⎛ 2m ⎞ 2 2 ⎡ 2m a ⎤
( )
2
4m a
0=⎜1− − ⎢ ρ +a + ⎥ dϕ − c dt dϕ
2 2
(107) ⎟ c dt
⎝ ρ ⎠ ⎣ ρ ⎦ ρ

The speed of light then takes two different values, depending on the direction
of emission:

(108) vϕ = c ⎡⎢ 2 m a ± 18m 2a 2 + 2m ρ 3 − ρ 2 a 2 − ρ 4 ⎤⎥
⎣ ⎦

This phenomenon is classically interpreted as a phenomenon of dragging of the

coordinate system (frame-dragging) and is not without evoking the idea of Ernst
Mach according to which matter and space would be closely linked. These two
values of the photon velocity is related to the presence of a cross term in dϕ dt
The presence in the stationary solution of such a cross term was envisaged in
1916 by the Dutchman J. Droste13 [21], but immediately rejected by this author as
unphysical.
Let us digress for a moment. Droste, a student of Lorentz, presented this work
on May 27, 1916, he was thirty years old.

13 https://fr.wikipedia.org/wiki/Johannes_Droste
 64

Fig.36 : Johannes Droste (1886-1963)

He starts with the same assumptions as Schwarzschild, and here is his result:

Fig.37 : Droste metric .

The result is absolutely identical to Schwarzchild's. In this very complete article,
everything is explained. The expressions of the geodesics and the rest. Later
Droste will say that when he presented his paper, he did not know that
Schwarzschild had just solved this problem three months earlier.
But let us return to Roy Kerr's work:

Since we admit in his solution the presence of an azimuthal frame-dragging,

what do we obtain if we consider in a solution invariant by action of O(3) a radial
frame-dragging. This results in the presence of a cross term in dRdt.
Keeping the Schwarzschid notations, this corresponds to the Eddington- metric,
which is deduced from the Schwarzschild metric by the change of variable

Rs R
(109) t = t' + δ Ln ( −1) δ = ±1
c Rs

Which gives:
(110)

⎛ R ⎞ ⎛ R ⎞ R
ds 2 = ⎜ 1 − s ⎟ c 2 dt' 2 − ⎜ 1 + s ⎟ dR 2 − R 2 ( dθ2 + sin 2 dϕ 2 ) + 2δ s c dt' dR
⎝ R⎠ ⎝ R⎠ R
 65

This situation was recently studied in [22]. The transit time along a radial
trajectory becomes:
(111) ν = − 1 : centripetal trajectory; ν = 1 : centrifugal path.

Rs h2 h2 Rs
λ R − δ ν Rs λ2 − 1 + − +
dt' 1 R R2 R3 δ = ±1
=
dR c R h 2
h2 Rs ν = ±1
ν( R − Rs ) λ2 − 1 + s − 2 +
R R R3

For radial paths ( h = 0 )

(112)

Rs
λ R − δ ν Rs λ2 − 1 + δ = ±1
dt' 1 R
=
dR c R ν = ±1
ν( R − Rs ) λ2 − 1 + s
R

Quand R tend vers Rs , cette contribution du temps devient infini if δ ν < 0

In the conditions:

dt' ν R − δ ν Rs
(113) !
dR c R − Rs

It is therefore possible to couple two metric solutions, two sheets connecting

according to the sphere of groove, playing then the role of one-way membrane,
surface that the masses can cross only in one direction. Let's consider the
following pair of metrics. In the first one, the masses can only enter the throat
sphere in a short time, but can only emerge in an infinite time, which is
equivalent to an impossibility. The opposite situation with respect to the second
layer, defined by the metric (91). Globally the transit, with entry into the first
layer and emergence into the second, takes place in a finite time. The reverse
transit is impossible.

⎛ R ⎞ ⎛ R ⎞ R
(114) ds 2 = ⎜ 1 − s ⎟ c 2 dt' 2 − ⎜ 1 + s ⎟ dR 2 − R 2 ( dθ2 + sin 2 dϕ 2 ) − 2 s c dt' dR
⎝ R⎠ ⎝ R⎠ R

⎛ R ⎞ ⎛ R ⎞ R
(115) ds 2 = ⎜ 1 − s ⎟ c 2 dt' 2 − ⎜ 1 + s ⎟ dR 2 − R 2 ( dθ2 + sin 2 dϕ 2 ) + 2 s c dt' dR
⎝ R⎠ ⎝ R⎠ R

A star in implosion would thus see its mass transferred into a second sheet,
according to a finite time.
 66

Epilog.

Frankfurt is the birthplace of Karl Schwarzschild. Every year a "Schwarzschild

Colloquium" is held at the Advanced Studies Institute in Frankfurt, devoted to
questions of cosmology and astrophysics. In 2017 the organizers of the
colloquium had invited the cosmologist Juan Malcadena, member of the the
Advanced Studies Institute in Princeton, USA. He began his conference, devoted
to the latest advances in the field of thermodynamics of black holes by saying :
- In 1916, when Karl Schwarzschild published his paper, the scientific
community had to spend some time before certain points were clarified. Today
these problems have been well mastered.

To show the way the community of specialists perceives these questions, since the
sixties, the simplest way is to reproduce the main stream interpretation as it is
presented in page 223 of the reference [18], and which translates a unanimous
consensus within the community of cosmologists of today and the partisans of the
Black Hole model.
 67

Fig.44 : Standard interpretation of the Schwarzschild metric

This present article shows that this view should be reconsidered.

Références :

[1] D.Hilbert. Die Grundlagen der Physik (Esrte Mitteilung). Nachrichten von der
Gesellschaft der Wissenschaften zu Göttingen. Mathematisch-ohysikalische
Klasse. (1915) 395-407.
English translation : The foundations of Physics (first communication). English
translation. Boston Series in Philosophy and Science. Copyright Springer, Vol.250
2007 ( 29$95 )
https://link.springer.com/chapter/10.1007/978-1-4020-4000-9_44
[2] D.Hilbert. Die Grundlagen der Physik (Esrte Mitteilung). Nachrochten von der
Gesellschaft der Wissenschaften zu Göttingen. Mathematisch-ohysikalische
Klasse. (1916)
English translation : The foundations of Physics (second communication). English
translation. Boston Series in Philosophy and Science. Copyright Springer, Vol.250
2007 ( 29$95 )
https://link.springer.com/chapter/10.1007%2F978-1-4020-4000-9_45
[3] Tilman Sauer : The Relativity Discovery : Hilbert’s First Note on Foundations
of Physis. See page 48. Correspondance Einstein-Hilbert au jour le jour. Page 48
https://arxiv.org/pdf/physics/9811050.pdf
[4] A. Einstein : Die Feldgleichungen des Gravitation ( The Field Equation of
Gravitation). Sitzungsber. XLIV , 8 , 778 and XLVI s. 799 (25 nov 1915)
[5] A.Einstein : Erlklärung der Perihelbeweggung des Merkur aus allgemeinen
Relarivitätstheorie. Königlish Preubische Akademie der Wissenschaften (Berlin) ?
Sitzungberichet (25 nov. 1915) p. 831-838

[6] K. Schwarzschild : Über das Gravitationsfeld Messenpunktes nach der

Einsteinschen Theorie. Sitzungsberichte der Königlich Preubischen Akademie der
Wissenschaften p. 189-196. jan 1916
English translation by S.Antoci and A.Loinger. On the gravitational field of a
mass point according to Einstein theory. arXiv :physics/9912033v1 [physics.hist-
ph] 16 dec 1999.
Traduction française à partir de la version allemande par H.Traccard :
http://www.jp-petit.org/papers/cosmo/1916-Schwarzschild-fr.pdf

[7] L.S.Abrams : Black Hole : the Legacy of Hilbert’s Error. Can. Jr. Phys. 67 , 9191
(1989). arXiv : gr-qc/01205055v1/13 feb 2001
 68

[8] A.Loinger : The Black holes do not exist « Also sprach Karl Schawarzschild »
(« Ainsi parlait Karl Schwarzschild »). arXiv :physics/9905030 may 1999nb
[9] S.Antoci and D.E. Liebscher : Reconsidering Schwarzschild original solution.
Astronomische Nachrichten 322 (2001), 3, 137-174
[10] S.Antoci and D.E. Liebscher : The Topology of Schwarzschild Original
Solution. arXiv : gr-qc/0308005v3 26 sept 2005.
[11] S.Antoci . David Hilbert and the origin of the Schwarzschild Solution. (2003)
https/arxiv.org/abs/physics/0310104
[12] A.A.Vankov : Explanation of the Perihelion of Mercury From General
Relativity ». Gen.Sc.Jr. Researchgate.net. (2021)

[13] J.P.Petit & G.D’Agostini : Cancellation of the singularity of the Schwarzschild

solution with natural mass inversion process. Mod. Phys. Lett. A vol. 30 n°9
(2015). arXiv : gr-qc/2103.12845/13 feb 2001( 23 March 2021)
[14] J.R.Oppenheimer and H.Snyder : On Continued Gravitational Contraction. Phys.
Rev. Vol. 56 pp 455-459 sept 1939
https://journals.aps.org/pr/pdf/10.1103/PhysRev.56.455
[15] M.D Kruskal : Maximal extension of Schwarzschild metric. Phys. Rev. 119, 1734
(1° sept 1060).

[16] R.C.Tolman : Static Solutions of Einstein’s field equation for a sphere of fluid.
Phys. Rev. 1939 nol.55 ( 15feb. 1939)
https://authors.library.caltech.edu/4362/1/TOLpr39.pdf
[17] K.Schwarzschild : Über das Gravitationsfeld einer Kugel aus incompressibler
Flüssigkeit nach der Einsteinschen Theorie. Sitzund der Phys. Math. Klasse V.24
24 Februar 1916
[18] R.Adler, M.Bazin, M.Schiffer : Introduction to General Relativity, Mc Graw Hill
Books Cie, 1975. http://www.jp-petit.org/books/asb.pdf
[19] H.Weyl : Zu Gravitationstherie. Annalen der physik, Vol.35 n° 18 (1917)
https://onlinelibrary.wiley.com/doi/epdf/10.1002/andp.19173591804 : 49 $
On the Theory of Gravitation. Reprinted in Gen. Relat. And Gravit. Vol. 44 pp 779-
810 (2012).
https://link.springer.com/article/10.1007%2Fs10714-011-1310-7 : 49 $

[20] R.P.Kerr (1963) Gravitational Field of a Spinning Body as an Example of

Argebraic Special Metrics. Phys. Rev.Letters 11 ; 237

[21] J Droste : The Field of a Single Centre in Einstein’s Theory of Gravitation and
the Motion of a Particle in that Field. Royal Netherland Academy of Arts and
Sciences (KNAW) Proc. 19 I , Amsterdam 1917 pp.197-215.
[22] P.Koiran : Infall Time in the Eddington-Finkelstein Metric, with Application
do Eddigton-Rosen Bridges. Inter. Jr. of Mod. Phys. D, 14 aug. 2021, to be
published. https://arxiv.org/abs/2110.05938
 69

[23] A.Eddington : A Comparizon of Withehead’sand Einstein’s formulae. Nature

113-192, 1924
[24] D.Finkelstein : Past-Future asymetry of the gravitational field of a point
particle. Phys. Rev. 110(4) : 956-967, 1958
[25] S.Bondarenko : Negative mass Scenario and Schwarzschild Spacetime in
General Relativity. Od. Phys. Lett. A Vol.34 n°11 (2019).
https://arxiv.org/pdf/1807.07412.pdf

View publication stats