Professional Documents
Culture Documents
Fieldsofforce 00 Bjeruoft
Fieldsofforce 00 Bjeruoft
"r,
CToIumlria Uniucvsitiji
in tht City of flexv ^oiii
1904
FIELDS OF FOKCE
A COURSE OF LECTURES
IN
MATHEMATICAL PHYSICS
1
DELIVERED DECEMBER
TO
23,
1905
9^
Nch) Yocfe
Agents
t;
->
I
On
the seventeenth clay of
fonr,
Edward Dean Adams, of New York, established in Colnmbia University "The Ernest Kempton Adams Fund for Physical Research" as a memorial to his son, Ernest Kempton Adams, who received the degrees of Electrical Engmeering in 1897 and Master of Arts in 1 898, and who devoted his life to scientific reThe income of this fund is, by the terms of the deed of search.
gift, to
to the publication
generous interpretation of
Adams and
of the Trus-
made
it
Kempton
Adams
P'und
Corporation
http://www.archive.org/details/fieldsofforceOObjeruoft
FIELDS OF FORCE.
ELEMENTARY INVESTIGATION OF THE GEOMETRIC PROPERTIES OF HYDRODYNAMIC FIELDS.
Introductory.
The
idea of electric
and magnetic
fields
duced by
Faraday
to
a distance.
took
place
among
yet if
And
we
what
is
an electric or a magnetic
field
of
force ? no one will be able to give a satisfactory answer. theories relating to these fields, but
We have
we have no
idea whatever of
intrinsically, nor
Above
all
other
fields
daily,
we have
My
should be happy
view of
it.
^yhat I wish to
tric
insist
upon
especially, is this.
Besides elecstrik-
and magnetic
which have
same claim
The
investigation of these
other fields side by side with the electric and magnetic fields will
per-
2
fectly plain
FIELDS OP FORCK.
and
anything
These other
in
fields
They
of dynamics.
For the
special case
when the
medium
is
fields
The
tiania.
results
originally
by
my
A. Bjerknes, of Chris-
But
much
facility.*
My
two
parts.
The
first
two
lec-
drodynamic
fields
the
for a perfect
The
electric field
may
simplest case by either of two vectors, the electric flux (or dis-
way
may
The
flux
and
the field intensity will differ from each other only by a constant
factor, the electric or the
supports the
field,
*For
v.
Bjerknes:
Wissenschuften zu
Christiania
am
April,
1903.
German
velopment
complete description of instruments and experiments, see v. Bjerknes, Vorlesimgen A. Bjerknei Theorie. Vols. I and II,
Leipzig, 1900-02.
6
liquid
On
may
field
of motion in any
moving
be described
by either of two
vectors,
which are
related to each other in the same way as the flux and the field The first of these intensity of electric or magnetic fields.
the velocity, and the other the product of the velocity As to its dynamical significance, this vector is into the doisity. momentum per unit volume or the specific momentum in the the
vectors
is
moving
fluid.
field intensity in
the
parison.
There
is
how
and
this
knowledge of the
2.
fields.
Geometric and
Dynamic
The
extent
of our knowledge of the different kinds of fields differs greatly. All the properties of the hydrodynamic fields follow directly from the
most trustworthy laws of nature, that is from the principle of the conservation of the mass, and from the principles of dynamics.
With
we
First,
we
field.
And
then, in a
more or
less superficial
we have
we
a very
the fields.
much
less
Taking the
therefore,
to
shall
be obliged,
of the
give to our
theory
a dualistic
form, comparing
dynamic properties
two
It may kinds of fields. the central point, where the geometry and the dynamics of the question are perfectly united, and thus make the comparison of
In
this lecture
we
fields.
4
3.
FIELDS OF FORCE.
Properties of the Field Vectors at a Surface of Separation.
As
how
hydrodynamic
well
known,
is
normal component
of the flux
On
any sur-
moving
liquid the
velocity
is
the flux.
which we consider excluded. From the continuity of the normal component of the velocity follows the discontinuity of the specific
momentum
fluid
suddenly
is
and the
specific
momentum
of the
Fur-
the induc-
Just as
tlie
the matter to the motion, the specific volume gives the measure of
the readiness of the matter to take motion.
The
specific
volume
thus
may
We
flux,
field intensity,
momentum
mobility
inductivity.
to
We
4.
have now
this
correspondence.
Charged
Particle
Expanding
or Contracting Particle.
first
lict us start
investi-
In the
field
and radially
in-
wards
It
if it
is
contained
field
produce a
of
charged particle.
It
will
specific
momentum
will de-
tluis
b,
a and
below).
electric
and a
is
radial
hydrodynamic
Therefore
for
The
ing, or of an to
always contracting
particle,
impossible.
make
moment, but
In
and a
in
For
if
contracting,
and vice
versa.
We
we
particles,
mean
state of
by a vector we call, by
we
call negative.
5.
Complex
Fields.
If
we
sufficiently small in
FIELDS OF FORCE.
tance between them, a field
is
produced which
fields.
is
The
may
The
of superposition.
known
curves running
of opposite
point between
same sign
which
suffi-
and 6 below).
if
we consider two
particles
produced by each
may be may
represented,
corresponding electric
fields.
And
this result
;
be transferred
motion
same phase, expanding simultaneously and contracting simultaneously, will give a field corresponding geometrically to that
combine the
fields
field
way we can
construct a corre-
sponding hydrodyuamic
fields
field,
An
6.
Oscillating Bodies.
What
we have said of electrified particles and the electric fields produced by them may be repeated for magnetic poles and the corresponding magnetic fields. But now the reservation must be made, that magnetic poles are in reality mere fictions. For a distribution of
state of intriuHic
real origin
of the
magnetic
with
field.
Such
met
in electricity.
Thus
poles,
sum of
ex-
But
also be
expansion or contraction at
which corresponds
;
to the
magnet.
a certain motion
it
will
ing particles are situated, and recede on that side where the contracting particles are situated.
face as a whole, directed
The
result
is
the contractions
is
zero, the
unchanged during
this motion.
Now
itself
remains unchanged.
We
can there-
with an incompressible
fluid
fluid,
masses a
We
a motion of in-
compressible fluid masses, produced by suitable forces, can be found, which will set up an exterior
field similar to that set
up by
sum
FIELDS OF FORCE.
a body in a state of
intrinsic polarization
therefore, a
moved through
by
suitable
We
state of motion.
The
is
The
case
when
it
is
That the heterogeneity has an influence upon the geometric configuration of the field, is obvious. For only when the fluid is perfectly homogeneous will there exist that perfect
separately.
symmetry
will arise.
ists
in the space
But
it
if
symmetry
is
is
lost,
and
is
On
well
dielectric has a
marked
field,
influence
upon the geometric configuration of the electric to the j)henomena of electrification by influence.
giving rise
Now,
two cases be
have
to
fluids,
of similar nature?
To examine
this question
to the
we
shall
dynamics of
Consider
vertical, containing a
filled
with water.
body and, apart from the body, completely The condition of equilibrium will depend upon
If the body
has exactly the density of the water, the buoyancy will balance the
it
will
remain
If
in
If
it
be lighter,
its
buoyancy
and
will
it
will tend to
less
move upwards.
weight, and
it
be heavier,
its
buoyancy
be
than
its
will tend to
move downwards.
heavy body
will
body
sink to the bottom, and the body of the same density as the water
will remain in
any
position.
buoyancy depends upon the action of gravity. But there exists a corresponding dynamic buoyancy, which is To do away with the influence of easily observed as follows:
This
static
gravity, lay the cylinders with their axes horizontal, and let the
cyl-
move suddenly
observed
1.
is
lighter
moved
its
cylinder,
a motion
is
moved
its
had a motion
until
it
The
and the
will
stops against
move backwards relatively to the cylinder, until the end. The effect is strikingly analogous to
this
the effect of statical buoyancy for the case of the cylinders with
vertical
axes,
and
10
free to
FIELDS OP FORCE.
move through
when they
are
is
nil
same density
as the water.
in the case when the body has the The body then follows the motion of
The motion of
the body
is
mass x acceleration.
As
the
And
this
law evidently
will
by suitable forces at moving water. For, the state of motion outside the body is then unchanged, and the pressure exerted by the water against any surface does not at all depend upon the conditions within the surface. Thus we find this general result, which is perfectly analogous to the Archimedian law
rest relatively to the
:
Any
mass
is
a fluid
kinetic;
buoyancy equal
product of the
mass of water
This law obviously gives also the value of the exterior force
in order to
make
the motion of the fluid, just as the Archimedian law gives the
is
This force
is
is nil, if
as the water,
it
is
lighter,
and
in
body
is
heavier. no such force act, we get the result, illustrated by the experiment, that the light body moves faster than the water and the heavy body slower, and thus, relatively, against the
And,
water.
11
in the
Analogous Hydrodynamic
Field.
From
and
tlie
princijile of
kinetic
buoyancy we
fluid,
geneous
The
magnetic
field.
For
we
have greater
electric, or
magnetic
To
let
attached
in a
below with a
similar
fine string.
And
in
another bottle
suspend
manner a
lead ball.
we shake the bottles, the celluloid oscillations, much greater than those of ball will remain almost at rest. With
If
the
oscillations than
medium.
And
body has
surrounding medium.
1
0.
The
is
obvious.
At
places of greater
and
widen.
They
will thus be
narrow
at places of great,
and wide
at
tiie electric
or
magnetic
field are
narrow
at places
small inductivity.
when
we can
easily prove
12
FIELDS OF FORCE.
We
suppose that
tlie
surrounding
fluid are
themselves
It
is
only in experiments
we must always
At
the
sides of the
action
and reaction.
it
From
But
we
two
Or, what
is
the
same thing,
product of the tnngential acceleration into the deiwity will have the
sides of the surface.
this result there can not at once be
From
components of the
or of the specific
momentum.
If,
which
moment no
longer be adjacent.
every particle has an invariable mean position, then adjacent particles will
components
Thus,
In
motion the
specific
is
13
already found (3), the same as for the electric or the mag-
As we have
We see
by the
electric or
the same.
The
and the
on
lines
And,
as
is
shown
in all treatises
angles of incidence and refraction are in the same ratio as the inductivities
In the hydrodynamic
on
same
from one medium into the other. This refraction of the hydrodynamic lines of flow according to
the
same law
magnetic
lines of force is a
in the
motion of super-
If I suddenly
move
along the rear wall of the glass, while the water sinks in front.
During the first instant of the motion, before we get the oscillations due to gravity, the law of the refraction of the tubes of flow is fulfilled at
Whatever be
and of
mercury
:
13.
We
get the
same law of
air,
of water and
ratio, 1
:
700.
The
quence of a law
electric or
strictly
14
1 1
FIELDS OF FORCE.
Experimental
Verifications.
We
of the results,
Some experiments
many,
These experiments were made with water motions of vibratory nature, produced
by pulsating or
field,
oscillating
bodies, using
produced
The Generator.
small air
valves.
simply use
drums,
and drawn
can work in either the same or in opposite phase, and so that the
pump
it
For convenience,
In Fig.
ditions.
is
springs,
to the pistons
wooden frame
in such a
c,
way
that each
is
free to turn
about a
horizontal axis,
Thus
either
pump
can be
15
The amplitude of
nent of the motion of the top of the spring along the axis of
cylinder
is
At 90
the amplitude
0, to
and the
The generator may be driven by a motor of suitable nature, As shown in the figure, we may use a
Fi(i. 1.
electric motor, or
for
motive power.
The use of the crank has the advantage that the amplitudes of the
oscillations of the springs are invariable
may
The
springs are
A
Two
A rubber
membrane
is
16
FIELDS OF FORCE.
that water admitted to the outer cylinder cannot pass into the
Under
suita-
membrane,
Fig.
2.
Fio.
3.
to the
pumps by
the connecting-rods.
The
17
charge-pipe.
An
electromagnetic vibrator
is
For
As
a pulsating body
we may
use
the
pumps
in
of the generator.
it is
the tube
end
a drum, which
membrane.
A
is
diagram
2.
shown
in Fig. 3.
The
oscil-
lating
body
ff,
made
b,
c,
in
two halves,
and attached
one
same material,
which reaches
connected with
h,
metal tube,
pump
and
The
rubber membrane,
rod,
_/",
e,
is
The sphere
is
made
as light as pos-
greater mass.
For convenience
of
oscillation the
in differis
Thus, two
oscil-
pumps
in the
Reyister'mg of
When
scribed,
it
cannot
indirectly
in
several ways.
For example, we can suspend small particles in the water and observe their motions, and we might even succeed
in getting
18
particles.
FIE1-D8
OF FORCE.
may
more mechanical method, depending upon the principle of kinetic buoyancy, is preferable. A body which is situated in the
oscillating masses of fluid will be subject to a periodic kinetic
buoyancy which
tion
tries to set
up
in it oscillations
as those of the
will,
water.
The amplitudes of
is
produced
may be
in-
creased by resonance.
The body
fixed
upon an
elastic wire,
and
it
The amplitude of
body
is
is
The body
the brush
is
made
One
When
may
The
and
be pressed
down and
is
the brush
marks an ink
to
line
upon
it.
registering device
then
moved
this
another place
in the fluid,
and
so on.
In
lines
Fields.
corresponding magnetic
with iron
filings.
fields,
known way
4, b,
sponding magnetic
bar magnet.
two bodies pulsating in the same phase. They represent the meeting of two radial currents issuing from two centers. Fig. 5,
b,
19
\'
'
<
mmmmiiw
'.'* 4 'ifk'
\-
It
MA w\:\
""-yvX-.
>-"
b
Fig.
4.
in opposite phase.
The diagram
gives the
20
FIELDS OF FORCE.
other.
Fig. 6,
h,
sign.
(
'
'
'^'-iWiii///
<^;^
^i;/^
'
Xs
^
1
'
'
'
'
'
'
'
'
III
' I
21
in the
and Fig.
7, 6,
tion of the
poles, of
which two have the same sign, and one the opposite.
/ /
XV
o
<-/"
/
by an
oscillating body,
and Fig.
8, h, the
corresponding lines
22
FIELDS OF FORCE.
- /.
"
i\\\
^^
23
geometry of
//^"
in a
surrounding homogeneous
for the case
24
FIELDS OF FOBCK.
tlic niediiini
when
fluid
surroundis
more
and
of other density.
to
easily seen,
by means of our
registering device,
and
to diverge
But
this registering
Experiments which we
have
On
We
have thus
magnetic
extent,
of datical phenomena.
And,
to
some
we have
by experiments.
It
is
of greater generality, or to
general nature?
tory remark
is
imjjortant.
befor
and magnetic
fields.
dynamic analogy
fields
to electromagnetic
time.*
The
now be
raised,
fields in
To
this
it
must be answered,
it is
25
it
As an obvious argument,
fields
may
be
fields,
The analogy
the restriction
to
magnetism
when
is
excluded.
where
others
isolated
exist.
To
this
argument
may
be added
of
But
fields
in spite
this, the
it
netic fields
makes
possible to formally
fields.
compare hydrodynamic
the following discusis
with magnetic
And
practical reasons.
much more
familiar
to us
iu spite
of the formal
The
lines of
force of these fields always pass through the magnets which produce
The magnetic
at all
lines
of force produced by electric currents, on the other hand, are generally closed in the exterior space,
through
To
produced by an
as
an
axis.
to include
we would have
It
is
to look for
hydrodynamic
fields
be pro-
duced
by
fluid
pressure
cylinder, for
instance,
making rotary
all
oscillations
around
no motion at
if
in
a perfect fluid.
it
true,
elasticity,
26
as does
FJELDS OF FORCE.
an aqueous solution of gelatine.
But, as we shall limit
we
17. Detached
to the Fields
Electromagnetism.
A
It
of Stationary
is
thus
made
impossible.
a very remarkable
fact,
however, that
But
we must give up
We
thus arrive at an
itself,
but
which
is
Fio.
9.
This analogy
is
that discovered
by v.
Helmholtz
According
in
his
to his
and the
fluid
field
in
is
which produces
it.
To
field
field
of one
this
by concentric
field
circles.
And
of a rectilinear current.
The hydrodynamie
27
and
this
is
shown
field
in Fig. 9,
strictly
of two rectilinear
same
direction.
dynamic
magnetic
field
of two
rectilinear
parallel vortices
it
which
have
is
strictly
analogous to the
At
the
in
the
field
Fio. 10.
it
will
The
field.
The
is
limited
itself,
apart
from
its
divergence
one limitation.
An-
28
otlier
FIELDS OF FORCE.
limitation
follows from
not vary in
Helmholtz's celebrated
intensity.
theorem,
that vortices
do
Therefore phenomena
Whichever view we take of the subject, the hydrodynamic analogies to electric and magnetic phenomena are thus limited in
extent.
To
will
it
seems necessary to
But we
We
nation of the
properties
whose geometric
we have
ir.
unknown
to us.
What we know
magnetic
forces
field is
this
upon by
the
field.
Under
may
phenomena
to,
Fakaday's
of a tension parallel
and a
Maxwell's
mathematical translation
of this idea,
is
merely hypothetical.
And
may
truth, investigators
have at
all
makthe
from which
all
dynamic
fields,
Maxwell
We
. .
have
supposed
counted
it
fi)r
how
it is
maintained.
make
29
30
FIELDS OF FORCE.
tliese
In spite of
all
Maxwell's
Max-
well
namic
wrote them.
it
so
order to
Tiie question
simply this
Consider an
electric, or
magnetic
field.
field
corresponding hydrodynamic
field,
magnetic fields?
Tliis question can
Oscillating Current.
Let
Body
in
a Synchronously
buoyancy pro-
it.
If
is
its
volume be constant, so
It will
But
sist in
if
placed by
mass of water
will
have a
maximum when
its
31
has
maximum in one direction, and a minimum when the acceleration its maximum in the opposite direction. As is seen at once,
mean value
zero in the period.
maximum
volume.
We
thus
to the action
is
that of the
its
when
the pulsating
body has
maximum
3.
volume.
Bodies.
case of
As
first
application of
tliis
we may consider the bodies. Each of them is in other, and we have only to
result,
examine the
ducing
it
Evi-
directed outwards
is
when
it
has
its
and
is
directed inwards
has
its
maxi-
mum
volume, and
Let us consider
first
same phase.
maximum
itself,
vol-
when the
maximum
volume.
;
The
minimum volume. And therefore one will have its maximum volume when the radial acceleration is directed outward from the other. The result, therefore, will be an apparent mutual repulsion. As the force is proportional to the acceleration in the radial current,and as the acceleration will decrease exactly as the velocity, proportionally to the inverse square of the distance, the force itself
will also
On
it
is
must
also be proportional to
two param-
32
eters,
FIELDS OF FORCE.
which measure
in a
pulsation,"
we
law
is
nn apparent
is
phase there
an
to the
product of the
square of
to the inverse
distance.
4.
Discussion. We have thus deduced from the principle of dynamic buoyancy, tiiat is from our knowledge of the dynamics of the hydrodynamic field, that there will be a force whicii moves
field,
unknown
to us, a force
wiiich
through the
electric field.
And
the analogy
For the law enunciated above has exactly the form of Coulomb's law for the action between two
mere existence of the
electrically
force.
the
opposite to that
field.
For
opposite electricities.
For bodies pulsating in the same phase produce a field of the same geometrical configuration as bodies charged with and bodies pulsating in the same electricity (Fig. 5, a and b) opposite phase produce the same field as bodies charged with
analogy.
;
b).
is
complete analogy
it
most aston-
ishing.
in the present
We
know very
well
why
the
in
force in the
this is
must have the direction indicated But a simple consequence of the dynamics of the fluid.
field
hydrodynamic
magnetic
field
we cannot
tell
it
at all
why
has,
33
we
find
preceding lec:
This
may
be obtained
as a
in different
ways.
pendulum by a
long india-rubber tube through which the air from the generator
is
brought.
Or
it
may
made of
brings the air from the generator and at the same time serves as a
torsion wire.
The
air
which
is
fixed in a support.
b,
The
air
channel
At
and
is
is
splierical
The movable
part of the
This
d, the little
cylinder
c,
through
air
channel continues.
The upper surface of this The never touch each other, but by adjustment
may
To
By
body
is
free to
move on
a spherical
34
FIKLDH OF FORCE.
surface with the pivot as center, and the eqnilibrium will be neutral
for a horizontal motion,
6.
and
Experiments
toith
Puhatinf/ Bodies.
Having
one pulsator
^tfi^ffrut>>tttrftni
yii,iii>ininn,i\
Fig. 11.
we
the two pulsating bodies attract each other (Fig. 12, a).
attraction
is
This
and
it is
Fig. 12.
The moment
6).
pulsations
is
With
it
may
35
be shown with tolerable accuracy, that the force varies as the inverse square of the distance, and
the intensities of pulsation.
is
In this experiment the mean value only of the force and the
progressive motion produced by
it
are observed.
By
using very
nomenon
which the
is
possible.
It
is
is
not a
little
the observed
progressive motion.
7.
Action of an Oscillating
is
Two
field
Into the
Then,
we bring
into application
action between
magnetic pole
in
be changed,
if,
for the
an oscillating body.
two oppositely pulsating bodies, we substitute For both produce the same field, and the
body
will evidently
produced.
We
An
oscillating
body
will act
ele-
laic
of poles
This result
an oscillator
may be
verified at once
by experiment.
If we take
is
in the
it
we
(Fig. 13, a)
when
the oscillating
body
as
it
approaches
36
FIELDS OF FORCK.
is
it
is
expanding (Fig. 13, i), the attraction changes to repulsion. To show how the analogy to magnetism goes even into the
smallest details the oscillating body
may be
longation of the
arm
of"
axis
of oscillation
will then
is
The
pulsating body
move
and come
into equilibrium in a
Fio. 13.
lateral displacement is
is
brought
move
as the
pulsating body.
8,
If,
in the
preceding
experiment, we take the pulsating body in the hand and insert the
oscillating
body
in the balance,
we cannot conclude a
priori that
a force equal and opposite to that exerted by the oscillating body upon the pulsating body. The principle of equal action and reaction
is
common
actions at a distance
But
for these
37
is
To examine
The
kinetic
subject
of kinetic buoyancy.
will give
volume, which
is
between
two places
in
the same.
For
at both ends of
body
will
But
if it oscillates
somewhat
different in direction
and
intensity, these
two forces
will not
The
is
direction of the
always tangential
If the
field
be represented by these
and
if
known
at
body
at every point
of
its
path
may
be plotted, and
As
we
sults,
will
be sufficient to
we have
treme
to
Fig. 14.
values
Let, then, the continuous circle (Fig. 14) represent the oscillat-
ing body in one extreme position, and the dotted circle the same
body
in the other
let
jwrtional to the accelerations which the fluid has at these two places
at the corresponding times.
The composition of
these
two
alter-
Let us now
We
And we
we only
38
FIELDS OF FORCE.
two oppositely pulsating bodies will be acted upon by exactly the same average resultant force as the oscillating body. From the
results
An
action of
a force similar
to that
acting
in the in the magnetic field, the only difference being the difference pole-law. the forces which follows from the opposite signs
9.
ing
of Experimental Investigation of the Force exerted by a PulsatLet us now insert the Body upon an Oscillatiny Body.
in the direction
of
its free
If a pulsator be taken
in the
hand,
it
when
the
approach one pole of the oscillating pulsating body is body (Fig. 13, a), and repulsion if it is made to approach the And, as is evident from comparison with other pole (Fig. 13, b).
made
to
body
is al-
have ways opposite to that acting on the pulsating body. in the case of magnetism. equality of action and reaction, just as
We
followed further
if
the
The
oscillating
its
c).
body
displacement, so that
sating
Tim
Oscillating
Mutual Actions betrceen Bodies. The pulsator held in the hand may now
the be replaced by an oscillator, while the oscillator inserted in it is still free to move along its balance is left unchanged, so that
axis of oscillation.
We
may
first
hand
then correspond to that with magnets in longitudinal position. bodies attraction in the case, (Fig. 15, fi), when the oscillating
get
We
INVESTIGATION
are in opposite phase.
OI'
DYNAMICAL PROPERTIES.
in
39
which the
magnets have poles of the same sign turned towards each other.
If the oscillator held in the hand be turned around, so that the
in the
may be brought
which
it
oscil-
lates in
the direction
Fio. 15.
balance.
Then we
The
that
free to
may now
its oscillation is at
move.
ing them,
we
oscillate in the
oscillate in the
same
oppo-
40
force
is,
FIELDS or FORCE.
as usual, the reverse, the
magnets repelling
in the case of
If,
similar,
finally,
and attracting
in
we
ance with
1,5,
axis of oscillation
get the small
f), we again
displacement
descriljed
We
which are
is
all
distinguished by cer-
an
infinite
symmetric
}K)sitions.
In
all
of them
it is
easily
dis-
the
in the
corresponding positions.
11. Rotations of the
OsdUcding Body.
We
have considered
But
in
general the two forces acting at the two extreme positions also form
a couple, like the two forces acting on the two poles of a magnet.
The
first eifect
of this couple
is
But
is
if this
effect
its
reached, and as a
series of
consequence of
oscillations
its
inertia
it
will generally
go through a
will turn
about
If the phase of
body
around
exactly
phenomena
is
The
pulsating body
may now
41
we shall
is
The
position of equilibrium
c),
lating
body
drawn from
if it
center to
that of the
body
and
oscillates at
right
it
will
while the intermediate dissymmetric positions of the fixed oscillator give intermediate dissymmetric positions of equilibrium of
the
movable
oscillating body.
Vsl
-&;
\.
M
b
i)
@
c
Fig. 16.
same as
is
two
a consequence of the
mag-
experiment
a position
of unstable
versa.
equilibrium in the
We
oscillations.
we have
have
but which
another density than that of the fluid also exert a marked influfield,
of the electric
field.
This action of the bodies upon the geometfield is, in the case
rical configuration
of the
of electricity or mag-
42
FIELDS OF FOKCE.
netism, accompanied by a mechanical force exerted by the field upon the bodies. We shall see how it is in this respect in the
hydrodynaraic
field.
As we concluded from
which
is
a body of
tlie
same
density than the water will be brought into oscillation with smaller
From
this
we conclude
But
that during the oscillations the body of the same density as the
both the light and the heavy body will in the two extreme posi-
and
if these
Fio. 17.
it
will
The motion cannot therefore be strictly periodic. As a consequence of a feeble dissymmetry there will be superposed
tion.
upon the
is strictly
force
which produces
this progressive
motion
or electrification by influence,
As we have already
shown
in the
magnetic
field.
43
same
relation to the
geometry of the
and vary
in the
in inten-
from place
same law
is
two kinds
in
of
fields,
always opposite
positions of a light
a shows these forces in the two extreme body, which oscillates with greater amplitudes
1
than the
fluid,
and Fig.
the light
7,
dynamic
magnetic
field,
body
subject.
its
Fig. 18,
a,
two extreme
makes
relatively to the
Fig. 18.
fluid,
which
is
the oscillation
which brings
it
And,
force
as
is
which oppositely corresponds to the force netic body is subject in the magnetic field.
to
which a diamag-
The
well
known laws
and diamag-
once to the motion of the light and heavy bodies in the hydro-
dynamic
field.
these laws
is
that of
44
FIELDS OF FORCE.
connects the force with the absolute intensity,
field.
Faraday, which
tion of the force,
Remembering
that
the direction
we conclude
will
The
light
body
move in
body in
13. Attraction
Pulsating or
and Repulsion of Light and Heavy Bodies by a If the field be produced by Body. an
Oscillating
is
very
For
its
maximum
will
at the sur-
and
always decrease
lating body.
To make
floating
wax.
to a
body by a thread
which either
floating
slides with a
minimum
It
is
is itself
by corks
on the surface.
the light body should never be fastened directly to the sinker, but
by a thread of
freedom of motion.
it
On
is
seen at once
to be repelled.
If one
is
sufficiently
body may
also be observed.
If the pul-
sating body be brought near the heavy body, an attraction of similar intensity is observed.
In both cases
it is
decreases
much more
easily proved,
power of the
distance,
which
distance found for the action between a magnetic pole and a piece
of iron.
If for the pulsating body we substitute an oscillating body, the same attractions and repulsions are observed. Both poles of the oscillating body exert exactly the same attraction on the heavy body, and exactly the same repulsion on the light body, and even
45
the equatorial parts of the oscillating body exert the same attracting or repelling force, though to a less degree.
As
is
easily seen,
we have
magthe
As
force depending
oscillations, or
mag-
netizations, decreases
may be
And
these
For one of the simplest magnetic experiments we can take a strong and a weak magnet, one of which is freely suspended. At a distance, the poles of the same sign will repel each other. But
if
This
may
We
two
if
is
pulsators, giving
dif-
ferent amplitudes.
At a
It
Many
critical point
from attraction
to repulsion,
may
be made,
all
show-
way
forces.
Orientation of Cylindrical
testing a
Bodies.
to
The
most common
method of
ism
is
magnetism or diamagnet-
to
The
is
easily
made
46
FIELDS OF FORCE.
light cylinder
is
The
transverse,
Oscillating Bodies.
The
Upon by Two
More Pulsating
or
force exerted
of iron
is
lelogram-law from the two forces wiiich each magnet would exert
by
were removed.
For the
direction of the
magnets
is in
resultant of the two vectors which give the direction of this increase
It
is
there-
we
considered,
it
must be emphasized,
i.
e.,
as a
means of the
or
sum of two
more
In
this
have great interest here, because they are well suited to show
how
phenomena
We
i^e
below the
same
vertical symmetrically
held in
its
position of equilibrium
more strongly than before. But if the second pole be a north pole, the iron will seem to be repelled from the point where it had previously stable equilibrium. It will move out to some point on a
circle, the
diameter of which
is
about
-^^
47
made with
a piece of bis-
muth and
It
in
the central point between two poles of opposite sign, and would
seem
to be repelled
Bnt
it
if
and
would be drawn
to this
above.
On
it
would be repelled.
These peculiar phenomena are at once understood
we
re-
member
sign
is
has a minithis
mum
(Fig. 5,
brought between
Then
if
be repelled from, and the heavy body attracted to the central point
But
if
that the two bodies pulsate in the same phase, the light body will
all
whose diameter
bodies.
is
about
-^^
At all
will
be repelled.
The
to
heavy body, on the other hand, will be repelled from the center
circle,
Besides
magnets on a piece of
soft iron,
we have
also actions
between any
two pieces of
upon by a magnet.
This
is
upon
The
same
line of
48
FIKI.DS
OF FORCE.
and
tlierefore
mal
sign facing each other, and therefore the chains formed mutually
repel each other, so that they
vals.
It
is
worth mentioning
filings
be made with
poles of
t'lo. 19.
all
the filings are changed at the same time, the direction of the
forces between
them
will be unchanged.
duced
oscillations in the
hydrodynamic
field,
opixjsite in
every case.
The
normally
layers
empty
is
4y
to the
bismuth
filings.
For
fairly
practical
reasons,
it
is
])referable to use a
heavy powder,
nicely,
must be
homogeneous.
if
mon
red lead,
moved by washing.
above which
is
This
is
The powder
immediately arranges
itself
Fig. 20.
surfaces
same and
22 gives a
section
As
is
curves thus
50
It
is
FIKLIW OF KOKCE.
worth remarking that the dynamical principle which exis
Kundt's
clust-figiires
in gases.
Our
figures also
show a striking
marks formed
in the
And
dynamical principle developed here does not fully account for the
peculiarities of
these
it
especially
great dimensions,
is
Fig. 21.
The
fossil ripple
marks, which
known
fields
dynamic
18.
and Ekrtric
We
phenomena of
electrostatics or of
diflference
Our
investigation
of the
geometry of the
field
showed
us
51
we meet
with difficulties
beyond
to the
this point.
phenomena of electromagnetism therefore necessarily supposes a more or less complete modification of the views which have
led us to the discovery of the partial analogy already developed.
To
we can hardly do
to
better
partial
than
analogy which
although
it
Fio. 22.
As we remarked when we
there
is
fields,
magnetic
why
it
same
is
im-
possible.
But taking
analogy as
it is,
52
FIELDS OF FORCE.
Let us
lation
consider
itself at
rest.
The motion of
circufield
be perfectly sym-
The
Nor
it
will
if
common motion
to
of
translation
be communicated
the
tank and
the vortex.
Otherwise
this
would be possible
to discover
by an experiment of
But now
the tank only, while the vortex, or a rotating rigid cylinder substituted for the vortex, be held
still.
There
symmetry
added
to,
rotating cylinder
on one
side, the
on the other
side, subtracted
As we have
a neutralization of the
two motions.
The
cylinder, there-
fore, is
field,
in
the direction
in
which there
This corresponds
homogeneous magnetic
as before
;
field,
the homogeneous
field.
The rectilinear cylindrical vortex which we have considered may now be an element of any vortex. Therefore we may draw this
general conclusion
;
is
station-
ary in space,
will, in the
hydrodynamic
field,
be subject to a force
53
As
special consequences,
we deduce,
allel vortices
thus to currents of
same
As
is
essential.
we come
back
to the case
where the
rectilinear vortex
as the tank,
and
was no
for
The analogy
this restric-
stationary electromagnetisra.
Thus
two reasons
As we saw
in the investiga-
tion of the
Now we
mechanical forces
when
The analogy,
it
there-
a limited one
but even in
its
limited state
may
give ua
suggestions.
19. Experiments with Rotating Cylinders.
results developed
Simple
friction.
cases of the
may
By means
of turbines driven by
tation,
we may
set
which
in turn
consequence of
One
such
may
a pulsation- or oscillation-balance.
It
is
necessary, however, in
in the
fluid,
With
in
this instrument,
it is
easily
shown
54
FIELDS OF FORCE.
We
effects the
just as a cylinder
of
Fio. 23.
55
diamagnetivity
effects
is
Even
this geometric
analogy
it is
accompanied with an inverse dynamic easily seen that the rotating and the resting cylin-
analogy
der attract each other, just as a wire, carrying an electric current, and a diamagnetic body repel each other.
III.
C. A. Bjerknes' Problems
and Methods.
originally
All
the
phenomena
by elementary reasoning
through
mathematical
Bjerknes. AVhile searching phenomena of hydrodynamics which should have the appearat a distance,
ance of actions
The
later
him
to results
which he
verified
by a
series of experiments,
We apparently deviate
the mathematical tiieory.
tlie
to
real.
may
in
to
And
ceded
perhaps and of wliich the greater part was lost after the exact
We
may
method of
tiie
discoverer,
our
subject.
matical theory
Proceeding now to the matheThe Problem of Analogies. we shall also, in one sense, deviate considerably from
56
67
At
the begin-
ning the solution of the problem of spheres was certainly the most
natural
test, for this
way of submitting the vague anticipations to a was the time when the theory of the action
rigorous
at a dis-
tance was predominant, and the discovery of the simplest and most
striking instances of apparent actions at a distance
fascinating result for a
man opposing this theory to strive for. The doctrine of action at a distance
is
the
to
our aim
now
field
to
fields.
The
even
more
new
views.
Of course,
if
there exists a close analogy between hydrodynamic and electromagnetic fields, this analogy
namely
in
hydro-
on the other.
am going
to
may be developed directly from these two sets of equations. The method thus indicated is, indeed, perfectly plain and
There
fields,
is
no
of hydrodynamic
To
we have
firet to
fields.
Maxwell's Theory.
is
and to-morrow.
fields
contained in what
tell
generally called
This
What
Maxwell's
58
equations give a very
FIELDS OP FORCE.
full
To
this
geometric theory
is
much
fields.
less
devel-
Maxwell's
the
good
But
this
agreement ceases
when we pass
will
case.
ing media, and, as will be seen, the full discussion of the analogy
have found
to
be
my
safest guide
will
emphasize two
more
the impressed forces, which play a great part, from a certain point
But
Heaviside's
be
I shall use
my own
economize
shall
things
reflected
in
analogies
and
the
notation.
In thus outlining Maxwell's theory I wish to emphasize that do not introduce anything new. What I introduce I have found in other authors, who were perfectly uninfluenced by the search
I
for the
hydrodynamic analogy.
The guarantee
for an unpreju-
Induclivity.
To
is,
a material
medium we
attribute
two con-
stants, defining
its
two kinds
of
fields.
of the
medium
/3.
and may be
ductivity,
in-
do not know the exact nature of the properties defined by They can, therefore, not be determined in absothese constants.
* Oliver Heaviside, Electromagnetic Theory.
Vol.
I.
We
London, 1893.
09
What we
/3
If a^ and
called
/S^
the
specific
and magnetic
inductivities
respectively,
of
the
medium which
has
the
and
/S.
When we
stant only,
medium
in rela-
by one
electric
we
tropic substances,
is
hetero-
geneity
may
be allowed.
to the fields
Electric
and Magnetic
We
To
give
and
been introduced.
states exist-
in,
medium which
states or processes
is
unknown
in
What
still
is,
we
shall
completely
show more by
the
common dynamical
in absolute
measure.
But
their
When
once
new law of nature allows us to independent equation containing the same unknown
netic vectors, and submit
write another
quantities,
we
them
to absolute
measurements
Provisionally,
we can only do
exactly
60
FIELDS OF FORCE.
more
unknowns than
relative
known
for the
quantities as if
it
were known.
But we
retain the
symbols
unknown
is,
why
fields is split
two
different,
first,
where the
vectors
and space
is
considered
fields,
where
imperfectly solved.
6.
Clasmficatton of the
Vectors.
The
vectors introduced
to
may
which they
relate.
On
As
forces
unA fluxes.
and
flu.ves.
:
Between
intensities
this re-
lation
by the multiplication of a
field
by the induc-
tivity of the
medium
a corresponding flux
formed.
They cannot
This
is
an
important remark.
much
geneous quantities.
But
field
in the case
of electricity, as well as in
intensities
may
61
Taking now another point of view, we can divide tlie field intensities into induced, and impressed or energetic fe/d intensities, and
the fluxes likewise into induced, and impressed or energetic fluxes.
fluxes
and
field intensities
we have
to
some
Maxwell's
But in order to complete the system formally, the impressed or energetic fluxes and field intensities are introduced. They represent certain states, or processes, under
certain circumstances existing in, or going on in, the matter,
and which are ultimately the origin of every electric or magnetic phenomenon. The intrinsic polarization in the permanent magnet,
or in the pyroelectric crystal,
is
therefore represented
by vectors by contact-
of this
class.
They
As
the
related
have termed them energetic vectors, a name given originally by C. A. Bjerknes to the corresponding hydrodynamic vectors.
From
gives the
or actual
field intensities,
We
have thus
in each
But
groilp of six vectors only two are really independent of each other,
magnetic
field.
Which
But a
pair of vectors
it
will
be
most convenient
this
fore
and the induced flMd inlensitij. These we theredistinguish beyond the others, calling them simply the flux,
every case when their qualities as actual
field
and
flux
and induced
62
7.
FIEI^DS
Notation.
OF FORCE.
It
is
at once evident to
which
class or
To
and
field intensities
On
the other hand, actual, induced, and energetic vectors are desig-
nated
l)y
the subscripts n,
i, e,
field intensity,
are distinguished
by the omission of
first,
and
the
first
and second
letter
is
Electric.
Magnetic.
Flux.
Field intensity.
Flux.
Field iutenaity^.
(^)
Actual
A
A, A,
B
B,
ba
Induced
Energetic
a
ae
1
b
b.
B'
Electric inductivity
tx.
Magnetic inductivity
ft.
what
namely
A=
a
A;
+A
a,,
B
b
,
= B, + B^,
b
()
=a+ A. =
fxa.,
=b+ B, = /3b,
B,=
/3b.
A =
By
and
aa
different eliminations
we can of
When we
intensity as
we need
the
have to be introduced.
As we
63
we
shall usually
have
to
employ
A
am
= aa + A^,
/3b
+ B..
But
it
may be
objectionable.
will
serve for
my
The
comparative
Time of Relaxation. Besides their electric and magnetic inductivities, some or most media have still an in8.
Conductivity,
The
constant best
is
first
by E. Cohn.
left to
itself,
If an elecits
in
a conducting
medium be
electric
The time
in
which the
fraction
during
value
(e
1/e of
is
initial
logarithms)
This
is
stant of the
in absolute measure,
and therefore
ductivity 7, to which
by the equation
-".
(a)
T=
not
A
is
perfect
known. It is convenient, however, in order to obtain a symmetry of the formulae, to introduce symbols even for
iV)
T'
= l-
64
9.
field
FIELDS OF J'ORCE.
Integral
Form
of the
intensity, actual,
first
member
(a)
A=
eta
+A
A
field
intensity,
and
vice versa,
electric
field intensity, is
Heaviside's expression.
of these laws
case the
To
we medium To
In
coincides at the
we
Denoting by
r the radius
f/r
vector
the vector-
by
rfs
surface, the
two
circuital
laws
may be
(&)
The
first
equation
is
generally
called the electric current through the surface, the first represent-
ing the displacement-current in the widest sense of this word, and the second the conduction current.
being merely
fictitious.
may
propositions.
The
electric
65
of the surface.
negative line
integral of the
electric field
intensity
round
the
To
have
we
definition of
two im-
electric or the
mag-
netic
the
volume
integral
respectively o(
E or M within
the surtace.
by the equations
JEdr = jAds
jMdr^JBds,
dr being the element of volume.
is
Thus
are,
of masses was
we now
to
be an empirical
fact,
and
law of nature
to
which the
subject,
and which
But
and the
will
be convenient to retain
To
ditions
usually to be added.
Thus,
it
is
generally considered
/9,
time
9
T are known
The same
66
is
FIKLDS OF FORCE.
generally
made
and
and magnetic
densities.
wanted.
This
knowledge
is
For
0,
time
is infinite,
\jT=
much
a^, and ;9^, that the relaxationand that energetic vectors and electric or
exist, A^
= B^ = 0, E =^ 31=0.
These
Form
From
we
can,
The
()
A
The equations expressing
= aa + A
the two circuital laws
may be
written
= curl b, k = curl a,
C
electric
and the
expressions for
= dA + ^^ = '^^ +
which
curl (A
X V) X V)
(div
A)V
B)V
+ +
1
y,
A,
ih)
curl (B
(div
y^B,
being the velocity of the moving medium, and dfi't the local time
is
differentiator,
d
(^3)
.Z.
d
cT^
+ V.v.
67
The second
fictitious quantities,
and
T B,
form
^=div
A,
M= div B.
these fundamental relations
To
we add
=,
J,
= 0,
^,
^=0,
Jf=0,
A=0,
B,
{d)
/3=/3,
= 0,
= 0,
which are
1 1
.
Stationary State.
The
fields, as
this
every varia-
a magnetic
field
every variation
time of a magnetic
field
is
electric
phenomena.
in
To
consider stationary
fields,
that
is,
fields
medium be
at rest,
V=
0,
and
and
the time.
The
reduce to
^
/7I
A,
()
k=
j,B.
68
FIELDS OF FORCE.
equations
is
conduction-current, which
fictitious
The
tric fields
and vice re/wt. But utilizing we can now simply consider the
fields separately,
and thus
treat the
two stationary
without any
fields,
we have
A=
(6J
curl a
aa
+A
= k,
E,
c
div
A=
B,,
To
tions the
be added.
condition that
The now
written
= a>
0,
^=
B.
/3o,
A,=
ih)
B=0, k = 0,
= 0, M=0, c = 0,
contains one fictitious
field
Each of the two systems of equations quantity. The equations for the electric
tric field contain the density
fictitious.
tionary magnetic current density k, and the equations for the elec-
If, in
we
the
suppose
tiie
current density to
be everywhere
nil,
we
get
69
A=
(a,)
Ota
+ A,,
curl a
= divA =
0,
JS',
M,
iK)
= a^, A=0,
a
^=^0B.=
0,
jE=0,
These
static fields exist
M=0.
field
13.
A
we
research
relating to the
com-
fields.
As an
occupy us
in the
electric or the
The expression
()
^=
is
i/Aat?T,
all
extended to
space.
Now
in the case
the flux
A=
we have
aa,
and, therefore,
(c)
*^
= X 2a ^'^^"^ ^ X 2
a,7i'T.
70
FIELDS OP FORCE.
let
lis
Now
form
A=
expressing
G, as
is
it
av<^
curl G,
<f)
The
may
then be
written
<I>
= ifA
\74>cIt
To
avoid circumlocution
we
the field no real discontinuity, every apparent surface of discontinuity being in reality an extremely thin sheet, in
which the
Further,
we suppose
to well
known
I J<^ div
Ach
-f-
^Jg
curl
&dT.
Now
the
first
div
is
and not
to the
surround-
ing ether.
If
we
induced
and energetic
parts,
we
get
curl a
= curl a
By
-f-
curl a
curl
a repre-
analogy,
By
be represented, be represented,
electric cur-
same way
according to
rents.
Ampere's
theory,
by a distribution of
Now
the
curl
a can
And
it is
current
curl
a,,
l/oiA,
from which
71
Therefore,
will be sufficient to
apply the
second integral
From
namely,
the
sets
of equations
it
is
seen at
field,
(d),
namely,
curl
bdr,
made up of the
and the
by which, accord-
ing to
sented.
a,
Ampere's
Now
only exists in material bodies, not in the ether, and the quantities
div
B and
it
We
whether
15.
which apply
us consider
surrounding ether.
Let
where
field
now
field
in the case
A=
0,
no energetic
0.
= 0, and
We
shall then
have
div
A=
0, 7
curl a
rt
= 0,
7
field
disappears
cZ).
But accord-
sum of only
part of the
same energy, which is the A must disappear in every But when both the flux A and the energetic
the
is
and there
:;
72
will exist
FIELDS OF FORCE.
no
electric field at all.
Thus
and no
elec-
Ij
there exists
no true
no
energetic flux,
electric field.
If there
16.
exists
Unique Determinateness of
From
this
result a
fields,
at once.
a,
and
both subject to
E.
The
A=
curl a
aa
+ A^,
= k, div A=U,
field
i. e.,
A^,
A"
= A' .- A, a. a" =
a'
As
is
0.
and no true
electrification,
and
it
Thus
the fields A,
a,
and
A', a',
73
According
is
to
electric field
electrification,
of
and of magnetic
magnetic field
is
which
is
laid
down
They contain
in the
And
field
perfectly clear
if
we succeed
later in represent-
field,
field
But before we proceed to the investigation of the hydrodynamic we have to consider the dynamic properties of the electric and
field.
the magnetic
10
IV.
The Maxwell
To
is
equations
upon
meaning
ical
is
perfectly
unknown
to us.
afforded
The
be this
safest
way,
in
seems to
start with
is
believed to represent
field,
and bring
energy.
The
clear
;
is
thus perfectly
First
of
all,
there
is
whether
it
should be
the actual fluxes and field intensities or only the induced ones.
Fol-
field inten-
write the expression for the total energy of the electromagnetic field
(I)
>P
= J lA
a/Zr
+ flB
hch.
Here, the
first
amount of
he
electric,
and the
74
75
iield.
Localization
tiie
Starting with
we can of
this
expression for
and bringing
into application
course de-
strictly in
form of expression
We
are able
it,
when
But
to
problem before
us, it is
very desirable
acting against the bodies as a whole, but also the system of ele-
mentary
bodies,
forces,
Of these
is
elementary forces
limited
experimental
needed,
which
is
We
do not possess
but we admit
two
principles':
we suppose
that
it
is
amounts of energy, but also of a distribution of energy in space. That this should be so is, a priori, not at all clear. The universal principle of the conservation
of energy
relates only
to
amounts of energy.
And
in the
to energy,
of en-
and
ness.
utility.
But Assuming
still it
this,
may have a more or less limited usefulwe admit as a working hypothesis, that not only gives the total amount of electric
also the
distribution of this
energy
iAa+iBb.
76
FIELDS OF FORCE.
To
less de-
termined motion of the energy, which in connection with the transformations of the energy regulates the distribution of the electro-
To
this principle
previous one.
The
idea of a determinate
in abstract
dynamics seem
it
be
though
may seem
to
have
And
may
fall
Electric
and Magnetic
Activity.
To
these abstract
and
general principles
we have
to
add
more or
less
by experiment.
The
is
of energy
is
This
rate
its in-
it.
And
to in-
is
is
made
field intensity,
every
current.
Maxwelf/s
c
of a moving
medium
= curl b, k = curl a,
we can
at once find the rate at
which energy
is
For, multiplying
FIELDS.
77
get
we
a
left
+b
=a
curl b
curl a.
hand member gives the rate at which this energy is supThe discussion of the right hand member therefore will show how the energy supplied is stored, transformed, or moved to other places. In this discussion we shall follow the method indicated by Heaviside.* 4. Storage, Trawiformation, and Motion of the Energy. To examine the right hand member of the equation we express the
plied.
The
induced
field intensities,
a=a
The equation of
(a)
a^ c
a,
b=b
b.
b^
= a
curl b
b^
curl a
curl b
b curl
a.
For
(b)
we
write, according
to a well
known
vec-
tor formula,
a
first
curl b
b curl a
= div (a x
b).
In the
troduce for curl b the developed expression for the electric current,
(III., 10, b^).
Thus
(c)
acurlb
= athat
dA
^+
a, curl
(A x V)+a- Vdiv
A + ^a^- A.
Remembering
a,.
A=
a
aa,^,
we
find easily,
^^
-^--
= a ^f +
dsi"
\^
,
da
dt
dt
"
or finally
*0. Heaviside
magnetic
field.
:
da da 'af,)-Ja^^+af. dt^^'^^"^~i^"dt^'^"dt
On
78
FIEr>DS
OF FORCE.
Now we
have
b^
da
da
And
tion,
if
we suppose
that the
its
we have
dajdt =^ 0, and
And
therefore
a.
(c),
the vec-
(Ax
right
V)
=Ax V
div
[a,
x (A x V)]
In the
first
change the
hand term we interchange cross and dot, and In the second order of factors by cyclic permutation.
triple vector
known formula
(c,)
a,.
we have then
curl(A x
(c.)
V)= V
and
(curl
aj x A+div
we
get
[(a,
A)V-(a. V)A]
Substituting
(c^) in (c)
a,
curl
^,^ (
a .)
y,A a
+V
(div
A)a
(a,
- i<va +
(curl
aJ X A]
+ div
{(A
full
aJV -
V)A}
curl a
= I (iB
(curl
bj
y, B b
+V
(div B)b
- ib^V/8 -H
bJ X B}
(b
VjB}
FIELDS.
79
and
(e)
are
now introduced
in ().
we
get
a.
c+b.k=^{|Aa+iBb}
+ ^Aa+ ^,B
b
(/)
(div A)a.
A}
B}
+
which
div
[ax
b+KAa + BbJV}
+ KA-aJV we have
is
+div-{-(a. V)A
is
(b.
V)B + KB-bJV
},
The
first
member
gives, as
said, the
rate of supply of
member
after
this
The
first
term
^^-{iAa+iB.bJ
gives the part of the energy supplied which
electric
is
simply stored as
in the unit
volume.
y,Aa+ y,Bb
gives the part of the energy supplied which
is
wasted as heat,
fictitious
magnetic
V
As
of the moving
the equation
is
the
common dynamic
moving
particle.
80
FIELDS OF FORCE.
mechanical force de-
pending on the
electric field,
^^^
f,
= =
A,
B.
The
first
common
motion of energy,
first,
the Poynting-fl.ux
a X
b,
by the vector
|(A-a.
B.bJV,
which
is
the velocity.
Finally, the last term gives, according to the
tation, that part of the
common
interpre-
medium which
is
field,
the flux of
(a
V)A
-1
(A
aJV
- (b
V)B
^(B bJV,
For
this
may
given by
is
a(A N)
- (lA
aJN +
b .(B N)
- (p
bJN.
stress.
first
This stress
in the
splits
up
into an electric
and a magnetic
And,
case of
isotropy,
of these
81
and
This
is
seen
when
the unit
is
drawn
first
parallel to,
to the corre-
sponding
lines of force.
The theory thus developed may be given with somewhat greater generality and with greater care in the details. Thus the anisotropy of the medium, already existing, or produced as a conse-
and
in the values
On
the other
But
founded
If
we knew
and magnetic
vectors, should
?
we then
in the
developements above
meet no contradictions
This question
may be
difficult
to answer.
derive from
it,
solely
conservation of energy
must of course be
only say, that
it
true.
But
we can
is
we
possess.
The Forces in
What
particularly
As
(div B)b
- ^b;V/8
-f-
(curl
bj x
B.
is correct, this
This
is
expression should give the true distribution of the force acting upon
11
82
FIELDS OF FORCE.
the elements of volume, and not merely the true value of the resultant force
is
The
obvious.
The
first
and
is
The
The elementary
in the
force
force
tivity
v/3,
of the induc-
and which
is
When we
consider
a body as a whole,
tlie
medium.
nil in
It will point
outwards
but
its
if the
tion.
But the
force,
which
is
the square of the field intensity, will therefore have greater aver-
field
intensity, quite
direction.
e.,
move
And,
in the
same way,
the
tivity than
surrounding
field.
move in The
the direction of
expression thus
Faraday's
well
known
in the
magnetic
field.
The
third term
= curl b +
curl b
=c+c
where
is
and
c^
83
may
be represented.
Ampere's theory, the permanent magnetism The last term of the expression for the force
bj X B
the well
=c
X B
c,
X B,
where the
first
term
is
known
should be the
electric
same
6.
as the
force
current.
may be wrong.
;
But however
if
we
we
subject.
For
we come back
to the results of
the
observations
A
(a)
= /(div
B)b/ZT
is
- /ib^V/S^T + /(curl bj
By
BrZr,
gaseous or fluid dielectric of the constant inductivity ^^, which is itself not the seat of any magnetism 31, of any energetic magnetic
flux B^, or of
any
electric current
c.
To
avoid
mathe-
matical
prolixity
we suppose
that
which
the body.
Thus
body has
all the
properties of
84
the ether.
FIELDS OF FORCE.
By
this supposition,
we
when transformations of
pass from the above
By
we can
To
find
split
the
=b +
b,
and we get
(b)
= /(div
J
B)bfZT
b)
x Bch
+ J,
where
= J(div
B)b//T
b ) v/3rfT
)
x Bdr.
term.
/(div B)hdT
Substituting,
(i")
/(curl b) x Bdr.
/(b bj Vy8c?T
- /^b; v/3f?T. =
/3b
Introducing in the
of these integrals
^b,,
we
gel
which we have
to
remember
works only
first
of the upon the vector immediately preceding it. In the two integrals of the right hand member we join the scalar factor ^ with the vector b,, upon which v works, remembering 0b^ = B^.
integral
factors.
- fB\vdT = - fbB^vdr +
/(b h^)v^dT
- fl^vbldr.
FIELDS.
85
remembering that
of the body,
JbB,V(ZT
(6"),
J'(b bJVyStZr
+ /jbfv/3(^T.
we get simply
J
-/bB,V(?T.
F =/(div B)hdT
The
resultant force
forces, given
fj
by
(curl b)
B.
if
fictitious
the
mentary
forces.
But
if
trustworthy, the reverse might also be the case, or else none of them may give the true values of the elementary forces, while both of them give the true values of the resultant forces.
7.
In writing
the ex-
we have
and
By
may
is
But
as this
artificial quantities,
the possibility that the expressions under the integral signs represent the real elementary forces
is lost.
The
= y8b + B,.
86
FIELDS OF FORCE.
This has the form of the true equation of connection, except that
the constant inductivity
/S^,
of the ether
is
introduced instead of
is
known
artifice
of
were permanent.
for the resultant
To
force
(6, c)
we first remark that in the second integral of the expression we can write /3 /S^ instead of ;8. Performing the integration by parts throughout the whole volume of the body and remembering that ^ /3^ disappears at the surface of the body, we get
-/|bV/3<?T= _/ibV(/3-
/SJfZr
in the
in connection with
(/3
B^ =
/3|,)b
+ B^,
we
get
- f^h-S7l3dr-fbB^\7dT = jB,bV<^T.
The
substitution of this in (6, c) gives the following
more com-
(6)
= /(div =
B)b(/T
-I-
/b ,bVf/T
-|-
/(curl b) x Bdr.
(div
B)b
-I-
B^bV +
(curl b)
B.
FIELDS.
87
still
sponding
elementary force
According
to a well
known
vector formula,
follows.
jB^bvdT =
Transforming the
to a well
first
J"B^ Vbf?T
/(curl b) x B,.(h.
member according
that B^.
surface of
at the
fB^.hvdT
x B^dr.
(),
= ySg div b +
F
we
get
/3
/(div b)bcZT
/3
/(curl b) x hdr,
which
force.
is
It
expressed here by a
f^
fictitious
elementary force
/3
(div b)b
/3
(curl b)
b.
The divergence of
is
The
force
rounding medium.
8.
Remimi
lectures to
electric,
sum up
fields
Using
tion
of the
even the
or B^ (IV.,
7, a),
Of
for
we note
more
system (III.,
88
FIELDS OF FORCE.
Electric
Magnttic
A=
(^)
aa,
= aa + A^, = V + A,,
we have
Tlien
which express
the relation between the field intensity and the current density,
(5)
Finally,
curl a
= k,
curl b
c.
we have
electrification, or
of magnetism,
(C)
divA=^,
divB
Jlf.
To complete
namely,
we have
finally
of equations gives,
/3
= ^,
31=0,
= o, B. = 0.
c
in the sense
Our knowledge of
According
complete.
to the analysis of
the expression
t,
^
'^
,
f,
'^
f,, 2
= (div A)a Ja^va + (curl a) x A aA^V, = (div B)b - Wv^ + (curl b) x B - bB.v.
FIELDS.
89
knowledge
is
reduced to this
we
get on integrat-
ing any of these forces for a whole body the resultant force which
the
may
(^;)
f,^ 3
or
(^J
f., 4
12
V.
GEOMETRIC AND DYNAMIC PROPERTIES OF THE HYDRODYNAMIC FIELD. GENERAL DEMONSTRATION OF THE ANALOGY TO THE STATIONARY ELECTROMAGNETIC FIELDS.
1.
Preliviimiry
Remarks.
Our
preliminary
investigations,
based
on
we
are
now going
examine more
closely.
According
we have no reason to look for an analogy extending beyond the phenomena termed stationary. The main feature of the analogy
is
velocity,
intensity
specific
momentum,
inductivity
To
facilitate the
comparison of the
fields
electrical quantities.
tric
hydrodynamic quantities by the same letters as the corresponding The symmetry in the properties of the elecand magnetic
fields will
make
it
the comparison with the electric field to the comparison with the
magnetic
2.
field.
The
two; the scalar equation for the conservation of the mass, generally
called the equation of continuity,
a being the
specific
volume of the
be written
and
tion of continuity
may
(a)
ait='^''^90
91
The
first
member
is
is
the
The
equality
Now
moving
fluid
etjuation of
The
first
member
is
moving
These
vector
sum
to the pressure,
tiie
it is
differentiating
Eulerian
^'^-
expansion
(^)
dt
= dt +
They
contain the
properties of the fields viewed from one central point, from which
their geometric
will be
It
our problem to
tain geometric
order to be
able to carry out the comparison with those other fields which
we
know
tion
two separate
sides
92
3.
FIELDS OF FORCE.
Equation of Continuity Equation fo^- the Density of ElednThe equation of continuity has the form of one of the fication.
field.
To show
this
we have
letter
to (IV., 8, C),
div
A= E,
is
the equation which gives the density of electrification in the electric field.
4.
The
as
dynamic
to
original
electric field.
field.
To show
the
actual specific
(a)
momentum,
a,
A=
d&
1 da -dt-^adt^''
= ^-''i'^
(2, ),
or,
-^ +
In the
left
('liv
A)a.
=ff,
v;).
electrification, div
A.
in the other
get
93
to a
known
vector formula,
we have
(curl
aJ
x A
+
E^
(div A)a
=i-^p.
Now,
field,
aj x A
According to
be written
left
Aav
or finally,
= aaav =
iva;,
Aav
Substituting this above,
= V(iO - Kva.
+ (curl aJ
x A
we have
aa
qI
+ V (Jaa^) all
|a;va
(div A)a
= i-wp,
giving us
E^
We thus seem
field.
to
have
But we
still
have the geometry and the dynamics of the hydrodynamic field united in one set of equations. To make the first step towards
the separation of certain geometric and dynamic properties from
field as
the
we consider
sum sum
Let us
momentum,
as the
sum
a^,
thus
=a+
The
("^
v&
dt
dt
+ ^( J^*) -
i^!
V +
(curl
aJ X A
(div A)a
=f
vp.
94
FIELDS OF FORCE.
Now we
say
a, to
to
a condition.
this condition
shall satisfy
the equation
(&)
,-,t
= - v(p + JO.
The
(c)
=*~
('^^^
^)^
^^'^''
~ (""'"^ ^"^
fields,
''^
6.
We
field.
ha%'e thus
which superimposed
But the
dynamic equations.
curl
How(A)
at
geometric equation.
and
we
get
d
^5-
curla=
0.
ct
To complete
voHex
momentum
is
the dynamic
density
thus invariable at
k for
we
get
k,
sity.
which expresses the local conservation of the dynamic vortex denAs regards its form, this is the same equation which in
electric field intensity a
tween the
8,
B).
And,
as the conservation of
to the
is
local,
equation
field for
((/)
cor-
responds exactly
the cases
Fundamental Geometric Properties of the Hydrodynamic Field. in representing the hydrodynamic field
95
sum of two
partial fields.
Writing A^
(A)
A=
aa
--=
aa
A..
Then
(B)
while from the
ity
field
curl a
= k,
we
calculate the veloc-
(C)
divA=^.
geometric properties of the stationary electric
field.
To complete
the investiga-
we shall have to examine whether we can introduce conditions corresponding to the supplementary conditions (IV., 8, D). The introduction of conditions of this
nature for the fluid system evidently involves the distinction be-
fluid,
which we have
free ether.
to
com-
which we have
to
The
part of
we
(A)
where
to be
a^
is
constant,
= u>
require the fundamental fluid
we simply
homogeneous.
(A)
we
require
it
^=^,
to
be incompressible.
There
is
(A)
for, at
k=0,
every point of space the dynamic vortex has, according to
96
FIELDS OF FORCE.
We
are there-
fore free to impose tlie condition that in the parts of space occu-
pied by the fundamental fluid this coustant shall have the value
zero.
this restriction
we
To answer
this
we must
refer to the
(-D,),
dynamic equation
shall
(5, c).
On
we
have
\7a=
we
in the
fundamental
On
account of con-
dition (D^))
shall
have div
= 0, so that
Writing
to
=a
-f-
a^
(T)^),
just introduced,
we
find curl a^
curl
a^,
da
Furthermore, we are free to introduce the condition that the exterior force f shall be zero for every point in the
fundamental
fluid,
of
(curl
X A.
shall also
Now
if at
any point
in space
\=
dt
0,
we
have
d& =0
'
i.
e.,
field
intensity a^ unless
existed previously.
is
The same
will
be true
corresponding
field
intensity
a^.
Nothing prevents
97
(A)
always
fulfilled.
A.
the condition
(DJ
And,
as
we have
seen,
always be
fulfilled.
Summing up
-{D^)
we
find that
we have
is
I
I
field.
The fundamental
fluid has
may have
may have an
inductivity
varying from point to point and differing from that of the ether.
The fundamental
traction,
fluid
may
just
we have no
may exist
in material bodies.
The fundamental
fluid
may
just as the
may
exist
fluid
just as
may
Under
dynamic
scribed
electric or
magnetic
field
are de-
between the
two kinds of
9.
fields.
Dynamic
seen
13
Properties of the
Hydrodynamie
Field.
It
is
easily
that
98
analogy will be joined
pose the condition that
FIELDS OF FORCE.
to this
geometric analogy.
For
let
us im-
da
/.
e.,
momentum
to use
When
we
will
now have
f
if
(curl a)
x A,
e.,
specific
an exterior force
by
(b).
there must act upon the system whose distribution per unit volume is given According to the principle of equal action and reaction,
tions force
by the
fluid system.
The
fluid
(F,)
f,
= - (div A)a +
Ja;;v
- (curl aj
x A,
is
which,
in form, oppositely
exerted,
magnetic
field in
10. Second
Form
The
is
we
show that even other conditions may lead to an analogy, in which we do not arrive at Heaviside's, but at some one of the
we
will
We
start again
Now,
momentum
a^,
momentum
A =
aa
-|-A^.
99
we have
d&
1 dA ,^. ^^-+(divA)a+^^^-'
= f-Vi.
Introducing in the
first left
5a
,,.
1 <^A
\
or,
left
to the
vector formula,
/ X
5a
-^-1-
(c)
Aa V4-
(curia)
^. .^ A + (divA)a+
dA _f
=i
syp.
(6)
we
member
id)
W + ^(^*' + K-^)+l ^t +
-f-
(^^^
A)a
- iaV
S^p.
(curl a)
X A
aA^V = f
Now, we
(e)
aa =_ v(;. + ^aa^-|-A
find that the other vector A^
a),
and we
1
must
dA
if)
dt'^^~
(^^^
^^^
"^
2-aVa
- (curl a)
A+
aA,v.
100
FIELDS OF FORCE.
b)
Both equations are different from the corresponding equations (5, and (5, e). But, as is seen at once, the new equation for the in-
by
ric
(jB).
We
same
(C).
set
equations as before,
(A)
dently the same right as before to introduce the restrictive conditions (-D,),
{D^,
(^3).
we
are entitled
moving
did not
fluid
particle canuot
if this
exist previously.
now
fluid,
and not
to
the points in
this difference
The consequence of
Finally, we see that to this geometric analogy we can add a dynamic analogy. Requiring that the energetic velocity be con-
served individually,
we have
dA
we
find that
under
a)
x A
aA^V,
E.^).
We
fields.
101
field
to
condi-
to be fulfilled.
We
most fundamental.
we
for a
electric or
When we
therefore equivalent.
And from
in the
hydrodynamic
field
= (div
f^
A)a
A,av
(curl a)
x A,
a.
and
(E^)
= a^
(div a)a
a (curl a)
The
tric
fact,
elec-
laws does not necessarily imply also a real analogy between the
things to which they relate.
Maxwell
expressed
it
the
analogy of the relations of things does not necessarily imply an analogy of the things related.
The
magnetic formulse to an
VI.
of formulje which
we have
fields.
de-
domain of stationary
electric,
or stationary magnetic
But
we
There
is
no contradiction involved
in
these re-
If,
we
siiall
electric
or magnetic
systems there
The hydroits
dynamic system is configuration. But apparently, at least, the systems M^ith which we compare them are
formulte therefore gives only a
magnetic
corre-
at rest.
The
two kinds of
fields,
which
exist
To get an analogy, not only in formulae but in we must therefore introduce the condition that the
hydrodynamic system should appear stationary
can be done in two ways.
in
This
can be in a
steady state of motion, so that the bodies are limited by surfaces of invariable
fluid can
shapes and
position in
space.
Second, the
103
Skady
State of Motion.
The
first
analogy, in which
specific
we supposed momentum,
the local
which we
(a) also
which
is
But
field
in the case
of a steady
state of
is
fluid,
moves tangentially
bodies.
4.
to
the
stationary surface
As
a^,
the motion
= 0,
volume,
A=
well
circulation,
If,
which
is
only possible
is to
then, there
be any
fluid at all,
the
The
is
momentum by which
this
motion
The
corre-
stream-lines are
The
sponding electrodynamic
tangentially to the
is also 5.
field,
a well
known
This
exterior field
104
FIELDS OK FORCE.
in
The most
the
always
be
fulfilled
if
material,
and a
distribution
bodies.
The hydrodynamic
is
ductivity
zero mobility.
in
The
and
their
positions
space as a consequence
of an infinite
in
Now
the case
momentum.
We
finite distribution
momentum and
of
and
momentum
Other interior arrangements can also be conceived which proThe condition of infinite diamagduce the same exterior field.
netivity
may
exist if
bility
make botlies appear to be The corresponding hydrodynamic case will we abandon the infinite inertia as the cause of the immodynamic vortices
in the bodies,
of the bodies and also dispense with the creation of any gen-
eral distribution of
and
if
we
in-
troduce instead,
sjiecial
in
space.
This
dynamic
infinitely diamagnetic.
is
jKissible.
In bodies of any
in-
we can
set
up any
and
105
Cor-
we can give
to bodies
the
produced by external
6.
forces.
In
fields will
be perfect.
And
with
we have an
by which the field tends to promotions of the bodies, and which must be counterforces,
The
ders,
are those showing the attraction and the repulsion of rotating cylin-
and the
attraction of a non-rotating,
by
a rotating cylinder,
As
diamagnetic bodies,
it
for the
ing to
Ampere's
in
theory.
We
magnetism, but
to
refers only
infinitely
diamagnetic
is
material.
fore a
The hydrodynamic
representation of a magnet
there-
Such bodies
apparent actions at a distance upon each other, corresponding inversely to those exerted by permanent magnets which have the
peculiar property of being constructed of an infinitely diamagnetic
material.
in
Lord Kelvin
Motion.
The hypothesis of
of
Vibi-atory
the condition of
is
steady motion.
14
momentum
vibratory,
its
106
curl, if it
FIELDS OF FORCE.
has any, must also be vibratory.
that
dynamic vortex
is
density,
a constant at every
The dynamic
vortex therefore must be everywhere zero, and the equations expressing the geometric analogy reduce to
(a)
A^,
fluid,
a,
E=0,
field,
A=0.
so that the analogy will not
The
equations thus take the form of the equations for the static
magnetic
static fields.
To
we may
a or 10,
g).
For both
We
have to return
the
unrestricted
equation for the energetic motion, and the form which in this case
leads to the most general results
is
(10,
),
which according
to (a)
reduces to
1
(c)
(J
^' = f -
(div A)a
+ ia^Va +
any
single
aA V. moment during
it
is
valid for
the
We
another
Periodic Functions.
To describe
we
shall
employ
the diiferent particles of the fluid will not have vibratory motions
describe this
The motion of the fluid will have mode of an elastic system. To fundamental mode we use a periodic function,/, of the
f{t
period t
()
thus
^)=f{t)-
107
finite
first
The
limits,
order.
must be subject
to the following
0,
conditions
during a period
it
shall
1,
thus
/' + T
-J
1
/< + !
At)dt
= o,
{<^)
provided
it
be periodic.
Any
periodic function
may be made
to fulfil
of a constant factor.
conditions
(d)
is
An
f(t)=V2sm2',r(*^+h\
the conditions that the period
is
From
first
order and that the mean linear value of the function for
is
a period
zero, it is
by any
first
finite factor
order.
We
may
(e)
thus write
f;'nf{t)dt<S,
is
where n
first
order.
9.
To
now make
that
it is
The motion
K)
= K,Jlt), E=EJ\t),
108
FIELDS OF FORCE.
A,
where
tiie
time, but
As
to their absolute
mean
For, from
equation (8,
c),
we get
1
Z^' + T
E-=
The
(rt,)
Ehlt.
T Jt
may
;
have different
and
(rtj),
and the rule of signs may be expres.ed thus the quantities K,^ and E^ have respectively always the same sign as
and
E had
The
E^
signs which the different quantities A,, and E^^ have rdativdy to
each other.
With regard
to tiie motions
(8, c)
determined by
(),
we can conclude
of the function
which never exceed a certain small quantity of the first order. And in the same way we conclude, that the change of volume produced by the ])eriodic velocity of expansion and contraction never
exceeds a small quantity of the same order.
order,
we can
a^
where
it
According
to this,
it
is
and of the
momentum
a.
Doing
A=
a
A,/(),
(6^
= a,/(0.
109
(a)
the quanti-
ties
(c)
curl
= 0, div A = ^
a,
(d)
= 0,
E,
0.
If these equations be
equations and
The equations
(c),
satisfy,
They
give,
same analogy
is
to an invariable electro-
magnetic
single
field as
moment.
The
similarity
so great that
it
is
not even
To
;
it is
sufficient to
change the
if these
quan-
momanner the
geometric
mean
field
intensities
the
in the
hydrodynamic
field.
10,
to
Fi
nally,
we have
to substitute the
motion
tions.
and perform the integration over a period of the oscillaUsing the property (8, b) of the function f, we find that
hand member of the equation disappears. Designating by mean value of the exterior force f and using the property of the function _/', we find
the
left
f^ the
(8, e)
f,:.
110
FIELDS OF FORCE.
mean
force exerted
by
(div
)a
ia''
v +
a,A,,
v.
The
momentary analogy, except that the varying quantities are replaced by quantities independent of The similarity of the expressions makes it unnecessary the time.
sion for the force in the case of the
to use
We
the force
(div
A)a
+ |aVa +
aA^V,
a and momentary values of specific momentum and velocity, respectively or as the quantities which represent in the way indicated In one case we arrive at the mean intensities of these quantities.
and
as the
moment
Both analogies
We
this
result, that in
the case of
field
A=
(o)
A^,
curl a
= 0, div A = B,
B=0,
= a^,
A=0.
which
And
are given
(c)
by
^2
= (<iiv A)a
-|-
JaVa
-|-
aA,v.
Ill
the force
f^
an open question
whether
electric or
magnetic
field is
fundamental, or only a
which gives the right value of the resultant force upon the whole
body.
We
field
this
the vibratory
hydrodynamic
electrostatic or
magnetic
resultant
In the hydrodynamic
whose
upon finite
or magnetic fiM.
To show
only
to
and
oscillating bodies
constructed
On
the
we have consid-
But these
to
no
restrictive conditioils.
Nothing prevents
fluid
bodies the same motion as they would have if they were constructed of solid material.
The
reactions exerted
upon them by
We
have nothing
to
add
to the
ogy.
But, to
make
we may
make
a simple application of
law
112
FIELDS OF FORCK.
= fEdr,
JE'cIt,
we
F r =
r being the distance
"'
^ T~
2'
particles,
and
a^ the induc-
is
not capable
which
set
up the
field
in
the water.
He
same form
trified particles.
Let us imagine
that, as
from
]X)int to point in
the space
F/'e'.
He
then arrives
which
is
pulsating body
e.
He
has,
His ex-
this, that
there
is
a force acte'.
But he
e'.
But
in spite of this, he
may
He may
see,
and
make
some process going on, in this medium. In the electrical case we have no direct evidence that this hypothesis is correct, although thus far, the development of our knowledge of electricity makes it extremely probable that there must be some truth in it.
ing, or
But
in
we have
113
medium
fluid
exists
it is
an incompressible
fluid.
And
momentum
Thus we
get a
But now let our hydrodynamic investigator proceed still further. Let him conclude with Maxwell, that the attraction and repulsion between the pulsating bodies must depend upon a stress in the medium. Following Maxwell's developments
13.
Maxwell's
stresses,
with
But
wrong.
iiis
Max-
well's
pulsating body,
Maxwell's
fluid.
We
can-
Maxwell's
wrong for the electric field. But we have full may be wrong, even in this case. To return to the hydrodynamic case, it is
where the error comes
to
in
in.
Maxwell
But
task;
and second,
to
motions.
And
it
is
a stress of
much simpler
imagined by
Maxwell, which
two
effects.
When we
stresses
by him, which reduce in simple cases to Maxwell's stresses. We cannot test Heaviside's developments in the same way as Maxwell's.
For we have no
electromagnetic phe-
to
the
nomena of
15
when he
But the
fact
114
srtesses,
FIELDS OF FORCE.
even in Heaviside's theory, are introduced only to ex])lain
field,
tenance of the
field itself.
And
Thus
" Our attitude towards the general application of the special
stress theory obtained should, therefore,
form of the
be one of
scientific scepticism.
some anti-Maxwellians,
" It
in an
is
agation of disturbances?
elastic
The
it
stresses
transmission of
it
might be
that
is to
But
it
we
E and H
to be
stress
may,
magnetic
field
and
jxjlar-
ized bodies, just as the pressure in the fluid accounts for both the
field
and
I,
p. 87.
110.
VII.
GENERAL CONCLUSIONS.
Remarks on Methods of Research and of Instruction
IN
I.
Theoretical Physics.
We
have
in the pre-
From
we have
fluid.
extended
this
of motion in a perfect
And
this
So
far as the
analogy ex-
electromagnetic and
to the relation
;
be-
image
in
a mirror
in
every characteristic
the characteristic
difference
and right
are interchanged.
it
may
be,
common
its
image.
The discovery of
eral considerations,
this extraordinary
analogy gives
:
rise to sev-
been exhaustive?
and one of the first is this Has our research Are tlie phenomena investigated by us the
still
been exhaustive.
Even
There
are, indeed,
and hydrodynamic
fields
116
ties.
FIELDS OF FORCE.
And
if
we no longer
fluids,
may hope
2.
greater generality.
The
question
now
arises
Are not
we have found
it
so entirely depen-
when we pass
to other
we have
We
origin to a great
And,
material medium, and furnishes the same basis for a possible geo-
fields.
On
fields
had
is
This
forces
mathematical
is
in fluids.
But the
its
special
and
extent, will
media.
Thus
perfectly accurate, if
we except the
It will there-
to
examine whether
is
pre-
is widenet^I.
The Fields
in
To
examine
GENERAL CONCLUSIONS.
the chances of progress along this line
it
117
advantageous to
elastic
will be
Now
it is
extenelastic
medium with
fields
The
Maxwell
for optical
NEL and
tic ether,
his successors
from the hypothesis of the transverse elasand the equations which he had himself developed to
the simplest possible case.
Let
being
the constant
medium
is
'''U
(a)
-aF
/t
= '^V=Umay be
written
As a and
de
= v^a/iU.
we can now
introduce
On
the velocity
This member
may
at the
in a modified
form,
curl
()
^.
The
equation
may
then be written
'di
118
FIELDS OF FORCE.
= - curl - U,
,
^B
-,
c'U
curl
Dt
or,
if
a dl
^ = curl a A,
,
.
'
we introduce the
specific
momentum
Ota,
a according
to the
equation
(e)
A
aB
we have
(/)
-5<
=-""''
,
On
{9)
(rf)
in (c) gives
-St
= ^'""^ '**'^b
If we introtluce
{(j)
= tia^,
form
Thus we can
equations
^^=curlb,
an
^=_
where the vectors
tions
curl a,
and
a,
B and
A=
Ota,
where
/3
GENERAL CONCLUSIONS.
But
this
119
equations for a
system
is
is
the system of
Maxwell's
medium which
isotropic,
electrically
is
and which
magnetomotive
forces.
And we
A
a
electric flux
velocity
. .
.specific
momentum
tt,a?
B
b
a.
intensity
electric inductivity
volume
of elasticity
/8
magnetic inductivity
density^/coeflf".
As
is
well known,
we
geometric analogy.
We
have used
freedom
to choose a
form
The extent of this geometric analogy is very great even though we have avoided full generality by neglecting heterogeneities and intrinsic forces. For it extends now to that point where the crossing of electric and magnetic phenomena takes place, the point at
which the hydrodynamic analogy ceased.
4. Dynamien of the Field in the Transverse Elastic Medium. These well known developments, which lead to the geometric
elastic fields,
know from
the investigation of
which
is
what we
mediate.
The explanation of this apparent contradiction is, however, imAs we have remarked, the energetic force in the hydrofields
dynamic
as
it
originated
in
namic equation.
is
medium,
approximato the
member of
the equation
its
proper form.
120
FIELDS OF FORCE.
Let US repeat, therefore, the preceding development, but starting with the equation
1
(/A
dt
()
a
left
= '^^^^ +
its
*'
in
which the
member has
we have
which we sup-
is
may
be developed in exactly
We
may
A=
and then equation
da.
aa
A^,
(o) in the
1
form
~^j
Vihaa?
+A
fJ'A
a)
^^^'
(div A)a
JaVa,
f,
+
corresponding to (V, 10, d).
(curl a)
X A
aA^v = a^V^U +
is
As
the
medium
supposed homo-
^a
^^
+ V (ia' + A
may now
,,
.V + a)
1 dA^ ^ -^^
(curl a)
x A
aA,V = A'V'U +
This
be introduced in equation
(a),
f.
then separated into two equations, just as in the case of the cor-
We
tem of equations
(6)
^*=AtV^U-v(aA.+
1
Jaa^),
dA
-jf
(c)
= i + aA,v - (curl
a)
x A,
where the
first is
The
first
of these equations
differs
from equation
(a) only
by
GENERAL CONCLUSIONS.
121
If
we agree
we may
still
de-
^ = curl
as
dt
D,
= curl a,
aa
where now
A=
A^,
But
if
we proceed
we have,
besides
by equation
f
(c).
And
an exterior force
must
= aA^V (curl
a)
A.
order
resjwnding to
5.
A^,
curl a.
the analogy.
And
phenomof
to indicate
common
of laws,
the laws
is
of
the fields
fields
of force
taken in a suitably
common
will
And
investi-
122
FIELDS OF FORCE.
may
even to the discovery of the true nature of the electric or magnetic fields.
6.
But investigations of
begun.
this
just
And
if
we
In
this
we have not only the well known geometric analogy, but also a dynamic analogy to the electrodynamic field, at least so long as we confine our attention to the analogy between the formal laws of the phenomena, and not to the phenomena themselves. And this dynamic analogy has exactly the same inverse nature as in the case of the hydrodynamic field. But it should be emphasized that this dynamic analogy, in the form in which we have found it, has not the same degree of completeness as
the geometric analogy.
we have
medium
to be
Most But the great drawback is this the dynamics of the electromagnetic field relates to two classes of forces, the electric forces and the magnetic forces, while our analysis of
gap can be
:
the elastic field has led us to the discovery of only one class of
forces,
forces, ac-
to the
It is true that,
we can change
elastic field
The
but
which pre-
Final Renmrks on
the
It
is
too
From
GENERAL CONCLUSIONS.
seemed
to exist
123
no dynamic analogy at
all.
its
we found
tromagnetic
field.
But even
and the
elastic field ?
is
problem which
carried out,
it
whether the
even
if
answer turns out to be positive or negative. answer be negative, the investigation of the
not therefore be completed.
It
is
And
fields
the
of force will
not at
all
to
be expected that
spond
to that
When
to
in
another form
What
fields shall
give
Even when
the problem
is
common
of force
elastic
medium.
problem of fields
it.
of vast extent.
We
124
8.
FIELDS OF FORCE.
Kinetie Theories.
The
problem of
fields
of force iu this
been present
in
but
From
the very
first
of
human
speculations on the
phenomena
to
be
the
way to render the phenomena of nature human mind. I need only remind you of
intelligible
the efforts
DemokDes-
cartes.
physical
Or
may mention
for
phenomena,
instance
dynamic
gases of
theories.
Or
may
Maxwell's
completely.
is
left except the building stones themselves, the atoms, which, however, have remained to this day an indispensable idea to the
natural
philosopher.
fate.
vortices
fruitful
Newton's
For the fact that phenomena of radiation could be explained according to his principle immensely facilitated the interpretation of
the
new phenomena of
still
radiation, discovered in
in radioactive substances.
The theory
if it is
it is
of light of
Huyghens and
Fresnel is
unshaken,
undulation theory.
in its original
light on
open to doubt whether it still exists form as a theory which explains the phenomena of
But
dynamic
principles.
hardly be satisfactory
,
For a dynamic theory of light will before we have a dynamic theory of electro-
magnetism.
GENERAL CONCLUSIONS.
125
physicists
may
still
highly appreciated,
as, for
instance,
As
to
Principles
Reactions
the
Phenomenoloffical
against
it is
exaggerations
are
always wholesome.
On
In accordance with
I will not enter
upon the
of
But
it
research,
the
which
Mach
Vienna, preits
The
Mach
is,
that the
phenomena
dom
kinetic theory, or to
nomena.
we should avoid
are ultimately
the positive
phenomena
phenomena
of pure kinetics, but that we should also avoid the negative prejudice that
tiie
phenomena of nature
The
will
deny
phenomena
by
far
phenomena.
research.
research exceeds
Tiie reason
is
We
laid
are
all
kinetic machines.
down
in
126
FIELDS OF FORCE.
nerves as an inheritance from the accumulated dynamic experience of our ancestors, and has been further developed without
in the cradle.
first
motions
And
furthermore, while
we have
knowledge
of the fundamental principles of dynamics, we have at the same time an objective view of dynamic phenomena as of no other
physical phenomena, from the fact that
we have
the power of
our senses at the same time, while for other phenomena, such as
sound, light, or heat, we have only one special sense, and for
others such as electricity, magnetism, or radioactivity,
still
we have
no special senses at
all.
No
dynamics soon developed into the model science, from the formal
point of view the most perfect of physical sciences, and in this
respect second only to pure mathematics.
why
gives
why we may
But
if
ing can be gained by giving up advantages which, for subjective reasons at least, are combined with the kinetic direction of
research, whatever be the final objective result of these researches.
The
It
is
such a
way of conducting
the investigation of the relations between physics and kinetics, which we have tried to realize in these researches on fields of force.
The
is,
made
the
made
laws found
for the
phenomena.
GENERAL CONCLUSIONS.
127
far
And
this
comparative method
is
applicable
outside the
it
problem of
fields
of force.
Indeed,
is
the
as
Boltzmann, Helmholtz,
methods
And
no
WiLLARD
Gibbs.
left
from
at-
tempting
Here there can be no mistake in regard to the agreement of the hypotheses with the facts of nature, for nothing is assumed in that respect. The only error into which one can fall, is the want of agreement between the premises and the conclusions, and this, with care, one may hope, in the main, to avoid." His method is exactly the same as that which we have tried to employ, namely the impartial research of each branch of physics by itself, but with comparison of the resulting laws, and with the
mechanics.
greatest possible caution
drawn from the analogies and the contrasts presenting themselves. The method is that of comparative anatomy. Is it too sanguine
a hope, that this method will, sooner or
relations of the different physical
later,
11.
On
Value of the
in Theoretical Physics.
Comparative Method
fw
Instruction
com-
128
FIELDS OF FORCK.
its
vahie
The
results obtained
in
by
this
apparently
And
be carried
and similar
facilitate, to
Especially will
throw light upon obscure theories, such as those of the electromagnetic field,
by means of perfectly plain and comprehensible theories field, in which every step can
rigorous mathematical conclusions, by elementary in-
made by
And
yet, this
perhaps
less essential
in
comparison
will gain
similar methods
used,
value of the methods, aud give him independence of view for his
work as an investigator. The arrangement of instruction according to principles by which the analogies at our disposal are used as nmch as possible for the benefit of the student, is a problem which has its own charm, infuture
to enter
upon the
seems to
details of
my
experiments
But before
It
seems to be an obvious
reflec-
draw
attention
away from
GENERAL CONCLUSIONS.
practically useful points.
129
To
dynamics
is
useful if
it
Now
water
is
practically
practical
motions of this simple medium, and not to the abstract fluid sys-
tems considered by
us,
was of
this opinion
my
study of
But
entific friends who were interested in the dynamics of the ocean and the atmosphere, I happened to see that certain theorems which I had developed to investigate the motions of my abstract fluid system had immediate bearing upon the motion of these two
media.
And
the reason
why
To work
investigators
had
most general
case, as
compressible according to an
depended upon the pressure But these very suppositions precluded from consideration the primary causes of the motions in the atmosphere and the For these primary causes are just the difierences of density sea. which do not depend upon the pressure, but on other causes, such as differences of temperature and .salinity in the sea, and differences of temperature and humidity in the atmosphere. While the old theorems of the practical hydrodynamics did not allow us to take up from the beginning the discussions of the circulations of the atmosphere and the sea, the thoerems which I had
idealized law, so that the density
only.
developed for
my
there-
ter
the
17
130
FJEI.1>S
OK
FORC^P:.
human
And
this is
an isolated
fact.
in theoretical
and
practice.
me
to
exemplify
which I
hydrody-
namic
fields
of force
in the
APPENDIX.
Vector Notation axd Vector Formula..
A A
A,^,
A^ B^
is
desig-
nated by A.
vector with the rectangular components B^,
is
desig-
nated by B.
(7.,
C^, C^ is desig-
nated by C.
Vector
Sum.
The
A + B =C,
are represented by one vector equation,
(1)
A-f-B
called the vector
= C.
A
and B.
is
sum of
Senlar Product.
designated by
vectors
(2)
The
scalar quantity
A^B^
+ AB^ +
A^B,
is
A B
and B,
A B=A B
Vector Product.
+A B +AB.
The
Ay B
-A B =
z
C, I'
are represented
(3)
AX B=
vector C
is
C.
The
vectors
two
and B.
The
131
132
FIELDS OF FOKCE.
normal
to
is
and B, and
is
directed
tiie first
Change of
In
For
this triple
product
it is
easily
proved that
of the factors
,
is
allowable, thus
'
In a vector-product one factor itself may be a vector-product. Cartesian development easily gives the formula
(5)
Ax (Bx C)=-(AB)C-f-(AC)B.
da ^x = dx
da
^.=
da
A=
are represented by one vector equation,
(6)
A = va.
differentiating
The
symbol
v or
The
vector
or
va
of this increase.
The
vector
v^
is
the scalar quantity a (compare the classical expressions pressuregradient, temperature gradient, etc.).
APPENDIX.
133
Spherical Derivation of a Scalar Quantity. The sum of the second derivations of a scalar quantity may be called the spherical derivative of this quantity, and the operation of spherical derivation
may
be designated by
v^ thus
Divergence.
is
The
scalar quantity
dAJdx + BA
jdy
-|-
dAJdz
A, thus
Carl.
The
dA^ By
dA V
Bz
Q"'
=
C^,
BA ___
_JC
BA
.
Bz
Bx
Bx
define a vector C,
By
''
which is called the curl of the vector A, and the three scalar equations are represented by the one vector equation,
(9)
Sphei-ical Derivation of
curl
C.
Vector.
The
Bz^
Bx"
'^
Bf a/
"^
~
~ ~
"
B3?
"^
"^
az^"
y'
Bx"
"^
By-"
B^
"'
134
FIELDS OF FORCE.
is
(10)
v'A=C.
Linear Operations.
The
three equations,
"
dx
^
,
dy
^
,
'
dz
dB
dx
dB
"
5
dz
dy
may be
(11)
AvB = C.
three scalar equations,
The
"
dx
'
dx
'
dx
<t-V'.^.f
may be
(12)
= -.'.
ABv=C'.
is
the
(13)
A vB = AB V +
(curl
B) X A.
following formulae
The
:
(14)
div otA
o(
div
A 4- A V
APPENDIX.
( 1
135
5)
di V
(A X B)
= A
curl
B+ B
a X
curl A,
/3.
(Hi)
curl (a
V /S) = V
we
get
Integral Fonnulce.
If dr
/Af/r
and
(18)
(Is
we have
/curlAf?s
(Theorem of Stokes).
whose normal
is
we have
jA-rfs
= JdivAdT.
Integrating
get
we
(20)
Ja v/3f/r = Jvax
V/3-(/s.
we
get
(21)
Ja V
a-d-T
ds,
(22)
fA
first
is
curl
BdT
= fB
curl
Adr
- fAxBds.
If in the
or B,
disappear.
When
it is
the
volume
whole space,
make
volume, we
136
FIELDS OF FORCE.
(23)
(24)
/a V Bf/T = - /B
J'AB
div Ar/r,
(It.
ch
= -/BA V
we
get
dr.
JB
div
(h
= J'AB V d r /(curl
New
B) x
:
For further
Gibbs-Wil-
a o >>
University of Toronto
CD:
Library
O in
a*
u o
c
fi
1^
POCKET
Acme
I
e
Library Card Pocket
Pat. "Kef. Index File"
o 3
r.i^cr
ade by
LIBRARY BUREAU
T-