Electrodynamics

Abhirup Datta, Siddharth Savyasachi Malu, FRAS

and Narendranath Patra
Copyright © 2022 Siddharth Savyasachi Malu and Abhirup Datta

P UBLISHED BY P UBLISHER

BOOK - WEBSITE . COM

Licensed under the Creative Commons Attribution-NonCommercial 3.0 Unported License

(the “License”). You may not use this file except in compliance with the License. You may
obtain a copy of the License at http://creativecommons.org/licenses/by-nc/3.0.
Unless required by applicable law or agreed to in writing, software distributed under the
License is distributed on an “AS IS ” BASIS , WITHOUT WARRANTIES OR CONDITIONS OF
ANY KIND, either express or implied. See the License for the specific language governing
permissions and limitations under the License.

First printing, January 2023

 Contents

1 Mathematical Preliminaries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.1 Vector Analysis 5
1.1.1 Vector Addition, Subtraction and Multiplication . . . . . . . . . . . . . . . . . . . . . 5
1.1.2 Gradient of a Scalar Field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
1.1.3 Divergence of a Vector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
1.1.4 Curl of a vector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
1.2 Co–ordinate systems 12
1.2.1 Catesian co–ordinates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
1.2.2 Cylindrical co–ordinates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
1.2.3 Spherical Polar Co–ordinates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
1.3 Mathematical Symbols 15
1.3.1 Kronecker Delta and Levi–Civita Symbols . . . . . . . . . . . . . . . . . . . . . . . . . . 15
1.3.2 Uses of Kronecker Delta and Levi–Civita symbol . . . . . . . . . . . . . . . . . . . . 15
1.4 Dirac Delta Function 16

2 Electrostatics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
2.1 Conducting Sphere in a Uniform Electric Field 21
2.2 Green’s Function Method 25

3 Conductors, Dielectrics and Capacitance . . . . . . . . . . . . . 27

4 Magnetostatics and Magnetic Forces . . . . . . . . . . . . . . . . . . 29

4.1 Force between current carrying wires: Biot–Savart Law 29
4.2 Divergence and Curl of the Magnetic Field 29

5 Magnetic Materials and Inductance . . . . . . . . . . . . . . . . . . . . 33

6 Time Varying Fields and Maxwell’s Equations: Electrody-

namics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
6.1 Issue with Ampere’s Law 35
6.2 Statics −→ Dynamics 35
6.3 Non–Uniqueness of Scalar & Vector Potential 36

7 Electromagnetic Waves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39

A The Frobenius Method: Special Functions . . . . . . . . . . . . . . 41

Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
Books 43
Articles 43

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
 1. Mathematical Preliminaries

In the study of electromagnetism, or electrodynamics, it is important to understand and

apply concepts of vector analysis and calculus, and certain other mathematical tools. This
chapter presents a collection of mathematical notation, definitions, identities, theorems,
and transformations which that are critical in the study of electromagnetism. A brief
discussion accompanies some of the less familiar topics and only a few proofs are given in
detail. For more details and complete proofs, the student should consult books listed in
Sources, References, and Additional Reading at the end of the chapter. Appendix §at the
end of the book summarizes the properties of Legendre polynomials, spherical harmonics,
and Bessel functions.

1.1 Vector Analysis

In the sections below, we shall recollect the basics of vector algebra and define three
important vector operations.

1.1.1 Vector Addition, Subtraction and Multiplication

For any defined vector quantity one can perform basic mathematical operations like
Addition, Subtraction and Multiplication. However in case of vector quantity one needs to
take the direction into consideration. Any vector ⃗A in three dimensions can be represented
with initial point at the origin (O) of a rectangular coordinate system. Now, let (Ax , Ay ,
Az ) be the rectangular coordinates of the terminal point of vector ⃗A. Thus, to express the
directionality from the O to A the vector ⃗A is defined as
⃗A = Ax x̂ + Ay ŷ + Az ẑ (1.1)

here, x̂, ŷ, and ẑ are the unit vector along the x, y, and z axis of the Cartesian plane. It is
important to note that x̂, ŷ, and ẑ are not a variables. However while performing vector
6 Chapter 1. Mathematical Preliminaries

additions or subtraction one should treat them as one. The following are the properties of
the vector additions/subtraction.
• Vector addition is commutative: ⃗A + ⃗B = ⃗B + ⃗A
• Vector addition is associative: (⃗A + ⃗B) + C
⃗ = ⃗A + (⃗B) + C
⃗
⃗ ⃗
• Vector subtraction in form of vector addition: A − B = A + (−⃗B)
⃗
• Vector addition multiplied by a scalar number is distributive: k(⃗A + ⃗B) = k⃗A + k⃗B

Example 1.1: Find a vector ⃗A directed from (2,-4,1) to (0,-2,0) in cartesian coordinates
and find the unit vector along ⃗A.
Answer:
⃗A = (0 − 2)x̂ + {−2 − (−4)}ŷ + (0 − 1)ẑ = −2x̂ + 2ŷ − ẑ

∥⃗A∥2 = (−2)2 + (2)2 + (−1)2 = 9

⃗A 2 2 1
a⃗A = = − x̂ + ŷ − ẑ
∥⃗A∥ 3 3 3
Example 1.2: Assume a three way tug-of-war situation between team-1, team-2 and team-
3 wherein each rope is 120◦ apart. The force exerted√by team-1, team-2
√ and team-3
√ are
4 3 −3 3−4 3+4 3 −3 3+4
5 N,√10 N, and 8 N along the directions 5 aˆx + 5 aˆy , 10 aˆx − 10 aˆy , and 10 aˆx +
3−4 3
10 aˆy respectively. Then find the direction of the collective force.

Answer: Here the direction of the forces are very important as this is an example of
tug-of-war the force are moving along a line outwards, so are the unit vectors. Thus, the
resultant force will be -
Fresultant = Fteam−1 + Fteam−2 + Fteam−3
Now,
4 3
Fteam−1 = Force magnitude × Force direction = 5 × ( aˆx + aˆy ) = 4aˆx + 3aˆy
5 5
√ √
= (−3 3 − 4)aˆx√− (3 + 4 3)aˆy
Similarly, Fteam−2 √
−12 3+16
and Fteam−3 = 5 ˆx + 12−16
a√ 5
3
aˆy √
−27 3+16
Thus, Fresultant = 5 aˆx + 12−4
5
3
aˆy

If ⃗A and ⃗B are multiplied, then the result will be either a scalar or a vector. It depends
on how multiplications of those are made. The two types of vector multiplication are:
1.1 Vector Analysis 7

1. Scalar product (or dot product)

2. Vector product (or cross product)
Scalar product (or dot product): The dot or scalar product of two vectors ⃗A and
⃗B, denoted by [ ⃗A · ⃗B ( ⃗A dot ⃗B)], is defined as the product of the magnitudes of ⃗A and ⃗B
(denoted either as A or |A| ) and the cosine of the angle θ between them. In symbols,
⃗A · ⃗B = AB cosθ (1.2)

Note that ⃗A · ⃗B is a scalar and not a vector. For example force and displacement are two
vectors, but their product is the work done by the force due to the displacement and is a
scalar.

Example 1.3: Show that ⃗A = 4x̂ − 2ŷ − ẑ and ⃗B = x̂ + 4ŷ − 4ẑ are perpendicular.
Answer:
⃗A · ⃗B = (4)(1) + (−2)(4) + (−1)(−4) = 0

Vector product (or Cross product): The vector or cross product of two vectors ⃗A and
⃗B, denoted by [ ⃗A × ⃗B ( ⃗A cross ⃗B)], is defined by

⃗A × ⃗B = AB sinθ n̂ (1.3)

where θ represents the angle between the vectors, n̂ is the unit vector in the direction
normal to the plane of the two vectors. The magnitude of cross product is defined as
|⃗A × ⃗B| = AB sineθ and the unit vector is defined as n̂ = (⃗A × ⃗B)/|⃗A × ⃗B|. Here, ⃗A × ⃗B is
neatly written down in component form as
 
 x̂ ŷ ẑ 
⃗A × ⃗B = Ax Ay Az 
 
  (1.4)
Bx By Ay 

So, ⃗A × ⃗B = (Ay Bz − Az By )x̂ + (Az Bx − Ax Bz )ŷ + (Ax By − Ay Bx )ẑ

Note that unlike the dot product, the cross product is non-commutative and non-associative
i.e., ⃗A × ⃗B = −⃗B × ⃗A and (⃗A × ⃗B) × C
⃗ ̸= ⃗A × (⃗B × C)
⃗

Example 1.4: Given ⃗A = 2x̂ + 4ŷ and ⃗B = 6ŷ − 4ẑ, find the smaller angle between them
using (a) the cross product, (b) the dot product.
Answer:
(a)
 
x̂ ŷ ẑ 
⃗A × ⃗B = 2 4 0  = −16x̂ + 8ŷ + 12ẑ
 
 
0 6 −4

q
|⃗A|= (2)2 + (4)2 + (0)2 = 4.47
q
|⃗B|= (0)2 + (6)2 + (−4)2 = 7.21
8 Chapter 1. Mathematical Preliminaries
q
|⃗A × ⃗B|= (−16)2 + (8)2 + (12)2 = 21.54
Then, since |⃗A × ⃗B|= |⃗A||⃗B|sin θ ,
21.54
sin θ = = 0.668 or θ = 41.9◦
(4.47)(7.21)
(b)
⃗A · ⃗B = (2)(0) + (4)(6) + (0)(−4) = 24

⃗A.⃗B 24
cos θ = = = 0.745 or θ = 41.9◦
⃗ ⃗
|A.B| (4.47)(7.21)

Example 1.5: Given ⃗A = x̂ + ŷ , ⃗B = x̂ + 2ẑ and C

⃗ = 2ŷ + ẑ, find (⃗A × ⃗B) × C
⃗ and compare
⃗ ⃗ ⃗
with A × (B × C).
Answer:
 
x̂ ŷ ẑ 
⃗A.⃗B = 1 1 0 = 2x̂ − 2ŷ − ẑ
 
 
1 0 2

Then
 
x̂ ŷ ẑ 
(⃗A × ⃗B) × C
⃗ = 2 −2 −1 = −2ŷ + 4ẑ

 
0 2 1

A similar calculation gives ⃗A × (⃗B × C)

⃗ = 2x̂ − 2ŷ + 3ẑ. Thus the parentheses that indicate
which cross product is to be taken first are essential in the vector triple product.

Example 1.6: Using the vectors ⃗A, ⃗B and C

⃗ of Example 1.5, find ⃗A · ⃗B × C
⃗ and compare it
⃗ ⃗ ⃗
with A × B · C .
Answer: From problem 1.5, ⃗B × C⃗ = −4x̂ − ŷ + 2ẑ. Then,
⃗A · ⃗B × C
⃗ = (1)(−4) + (1)(−1) + (0)(2) = −5

Also from problem 1.5, ⃗A × ⃗B = 2x̂ − 2ŷ − ẑ. Then,

⃗A × ⃗B · C
⃗ = (2)(0) + (−2)(2) + (−1)(1) = −5

Parentheses are not needed in the scalar triple product since it has meaning only when
the cross product is taken first. In general, it can be shown that

⃗ ⃗y ⃗ z
 
A A A
 x
⃗A · ⃗B × C
⃗ = B

⃗ ⃗y
B ⃗ z
B
 x 
C⃗x ⃗y
C ⃗z
C
As long as the vectors appear in the same cyclic order the result is the same. The scalar
triple products not in this cyclic order have a change in sign.
 1.1 Vector Analysis 9

1.1.2 Gradient of a Scalar Field

For any scalar function f , the total differential can be written in cartesian coordinates as
∂f ∂f ∂f
df = dx + dy + dz (1.5)
∂x ∂y ∂z
Now, since dr = dxx̂ + dyŷ + dzẑ, we can rewrite the above total differential as a dot
product of dr and another vector in the following way
 
∂f ∂f ∂f
df = x̂ + ŷ + ẑ · (dxx̂ + dyŷ + dzẑ) (1.6)
∂x ∂y ∂z
In the above equation, the expression in the first bracket on the right hand side is known as
the gradient of the scalar field f ; in other words,
d f = ∇ f · dr (1.7)
where
∂f ∂f ∂f
∇f = x̂ + ŷ + ẑ (1.8)
∂x ∂y ∂z
Starting with the expression for the gradient of f , or ∇ f in cartesian coordinates, we can
derive the expression in different coordinate systems too.
However, let us first understand the physical meaning of the operation that we are
calling a gradient. Let us look at one and two dimensional interpretations first. In one
dimension, all we get is
∂f df
≡ (1.9)
∂x dx
which is just the differential coefficient or the slope of the function f . In two dimensions,
we get

1.1.3 Divergence of a Vector

In the previous section, we used the vector operator ∇ to define a vector by operating
it on a scalar. We can also do the reverse, i.e. operate ∇ on a vector field to produce a
scalar. Before proceeding any further, let us first look at the use of such an operation and
its physical meaning. Frequently in physics, we need to measure the flux of a flow going
through a surface – a simple example is the flux of water passing through the surface of
a net kept in flowing water, e.g. river. Other examples, most notably in electrodynamics,
also need this concept as we shall encounter in the following chapters.
Clearly, defining a flux as
ˆ
Flux = A · dS (1.10)
S
makes sense. However, this does not tell us whether an object is a source or sink of flux.
Therefore, it is usually more useful to define the flux through a closed surface in this way:
‹ ‹
Φ= A · dS ≡ A · n̂dS (1.11)

where n̂ is naturally the unit vector perpendicular to the surface. Note, however, that there
are two shortcomings of this way of defining a flux through a closed surface:
 10 Chapter 1. Mathematical Preliminaries

´
Figure 1.1: Calculating the integral A · dl in a closed loop in the y − z plane.

• There is no normalization, i.e. the area can be small or large

• We only get the flux over the entire surface, and won’t get any estimate of point-to-
point variations in the flux
Both these issues can be easily resolved in the following way. Note that any closed surface
naturally encloses a volume. We can therefore divide the flux through any closed surface
by the volume enclosing it, and this standardizes the flux value. If we now shrink the
volume and therefore the surface area to a point by taking a limit, then we address the
second issue as well. This is defined as the Divergence of the vector field:
˛
1
∇ · A = lim A · dS (1.12)
∆V →0 ∆V ∆V

In order to get an expression for the divergence of a vector in cartesian coordinates,

let us consider a parallelepiped as shown below, so that the largest area is parallel to the
y − −z plane.

1.1.4 Curl of a vector

The general definition of the curl of a vector follows from the Stokes’ Theorem:
˛
1
∇ × A = lim A · dl (1.13)
∆S→0 ∆S C

Technically, of course, this is not quite correct! ∇ × A is a vector quantity and the right
hand side is NOT a vector. We must, therefore, write three of these equations, one for each
1.1 Vector Analysis 11

component along the x, y and z axes, thus

˛
1
[∇ × A]x = lim A · dl
∆S→0 ∆Sx C
˛
1
[∇ × A]y = lim A · dl
∆S→0 ∆Sy C
˛
1
[∇ × A]z = lim A · dl (1.14)
∆S→0 ∆Sz C

We can evaluate the first of the above expressions, and calculate the x−component of the
curl. To set this up, refer to Fig.(1.1). Let ABCD be the closed curve of infinitesimal¸ dimen-
sions enclosing an infinitesimal area ∆Sx along which we will evaluate lim∆S→0 ∆S1 x C A·dl.
Thus, we get, taking a middle–value of the vector A for each segment,
ˆ
∆y
For AB : A · dl = A(y + , z)∆y
2
ˆ
∆z
For BC : A · dl = A(y + ∆y, z + )∆z
2
ˆ
∆y
For CD : A · dl = −A(y + , z + ∆z)∆y
2
ˆ
∆z
For DA : A · dl = −A(y, z + )∆z (1.15)
2
Adding the first and third lines, we get
 
∆y ∆y ∂ Ay (z)
AB and CD : A(y + , z) − A(y + , z + ∆z) ∆y = − ∆z∆y (1.16)
2 2 ∂z
Similarly, adding the second and fourth lines in Eqs.(1.15), we get
 
∆z ∆z ∂ Az (y)
BC and DA : A(y + ∆y, z + ) − A(y, z + ) = ∆y∆z (1.17)
2 2 ∂y
Noting that ∆Sx = ∆y∆z, we get, finally, for the curl of A,
˛
1 ∂ Az ∂ Ay
[∇ × A]x = lim A · dl = − (1.18)
∆S→0 ∆Sx C ∂y ∂z

Example 1.7: Compute∇ · ⃗F and ∇ × ⃗F for ⃗F = x2 y x̂ − z3 − 3x ŷ + 4y2 ẑ.

Answer: Let’s compute the divergence first and there isn’t much to do other than run
through the formula.
∂ 2
∂
∂
∇ · ⃗F = 3x − z3 + 4y2 = 2xy


x y +
∂x ∂y ∂z
 

 x̂ ŷ ẑ 

∂ ∂ ∂
∇ × ⃗F = 
 

∂x ∂y ∂z 
x y 3x − z 4y2
2 3
 
 
 12 Chapter 1. Mathematical Preliminaries

∂ ∂ 2
∂ ∂ 2
∂ ∂
4y2 x̂ + 3x − z3 ẑ − 4y2 ŷ − 3x − z3 x̂





= x y ŷ + x y ẑ −
∂y ∂z ∂x ∂y ∂x ∂z

= 8yx̂ + 3ẑ − x2 ẑ + 3z2 x̂

= 8y + 3z2 x̂ + 3 − x2 ẑ

x3 y2
Example 1.8: Compute ∇ · ⃗F and ∇ × ⃗F for ⃗F = 3x + 2z2 x̂ +


ŷ − (z − 7x) ẑ.
z
Answer: Let’s compute the divergence first.

x3 y2 2x3 y
 
∂
∂ ∂
∇ · ⃗F = ∇  ⃗F = 3x + 2z2 + + (7x − z) = 2 +
∂x ∂y z ∂z z

 

 x̂ ŷ ẑ 

 ∂ ∂ ∂ 
curl ⃗F = ∇ × ⃗F = 
 
∂x ∂y ∂z 

 x 3 y2 
 3x + 2z2 7x − z 
 z 

x3 y2 ∂ x3 y2
   
∂ ∂ ∂ ∂ ∂
3x + 2z2 ŷ+ 2



= (7x − z) x̂+ ẑ− 3x + 2z ẑ− (7x − z) ŷ− x̂
∂y ∂z ∂x z ∂y ∂x ∂z z
3x2 y2 x3 y2
= 4zŷ + ẑ − 7ŷ + 2 x̂
z z
x3 y2 3x2 y2
= x̂ + (4z − 7) ŷ + ẑ
z2 z

1.2 Co–ordinate systems

1.2.1 Catesian co–ordinates
The notation for cartesian co–ordinates for vectors is

A = Ax x̂ + Ay ŷ + Az ẑ. (1.19)

We always use r to denote the position vector and ri to indicate either x, y or z, so that

r = xx̂ + yŷ + zẑ. (1.20)

The gradient or ‘del’ operator is

∂ ∂ ∂
∇ = x̂ + yŷ + zẑ . (1.21)
∂x ∂y ∂z
 1.2 Co–ordinate systems 13

The expression for the ∇2 operator, operating on a scalar field, which we will discuss later
in the chapter, is particularly simple (this operator can operate on vectors as well and in
that case, the scalar can simply be replaced by the vector):

∂ 2ϕ ∂ 2ϕ ∂ 2ϕ
∇2 ϕ = + 2 + 2. (1.22)
∂ x2 ∂y ∂z

1.2.2 Cylindrical co–ordinates

Fig.(1.2.1), on the left panel, shows the description of the cylindrical coordinate system,
which has the coordinates (ρ, φ , z). The transformation of co-ordinates to the cartesian
system is

x = ρ cos φ y = ρ sin φ z = z (1.23)

The notation for the components and unit vectors in this system is

A = Aρ ρ̂ + Aφ ϕ̂ + Az ẑ. (1.24)

The volume element in cylindrical coordinates is d 3 r = ρdρdφ dz. The gradient or ‘del’
operator in this system is

∂ ϕ̂ ∂ ∂
∇ = ρ̂ + + ẑ . (1.25)
∂ρ ρ ∂φ ∂z

The divergence, curl and laplacian are, respectively,

1 ∂ (ρAρ ) 1 ∂ Aφ ∂ Az
∇·A = + + (1.26)
ρ ∂ρ ρ ∂φ ∂z
     
1 ∂ Az ∂ Aφ ∂ Aρ ∂ Az 1 ∂ (ρAφ ) ∂ Aρ
∇×A = − ρ̂ + − ϕ̂ + − ẑ (1.27)
ρ ∂φ ∂z ∂z ∂ρ ρ ∂ρ ∂φ
1 ∂ 2 Aφ ∂ 2 Az
 
2 1 ∂ ∂ Aρ
∇ A = ρ + 2 + . (1.28)
ρ ∂ρ ∂ρ ρ ∂φ2 ∂ z2
 14 Chapter 1. Mathematical Preliminaries

1.2.3 Spherical Polar Co–ordinates

Fig.(1.2.1), on the right panel, shows the description of the cylindrical coordinate system,
which has the coordinates (r, θ , φ ). The transformation of co-ordinates to the cartesian
system is
x = r sin θ cos φ y = r sin θ sin φ z = r cos θ (1.29)
The notation for the components and unit vectors in this system is
A = Ar r̂ + Aθ θ̂ + Aφ ϕ̂. (1.30)
The volume element in spherical polar coordinates is d 3 r = r2 sin θ drdθ dφ . The gradient
or ‘del’ operator in this system is
∂ θ̂ ∂ ϕ̂ ∂
∇ = r̂ + + . (1.31)
∂ r r ∂ θ r sin θ ∂ φ
The divergence, curl and laplacian are, respectively,
1 ∂ (r2 Ar ) 1 ∂ (sin θ Aθ ) 1 ∂ Aφ
∇·A = 2
+ + (1.32)
r ∂r r sin θ ∂θ r sin θ ∂ φ
     
1 ∂ (sin θ Aφ ) ∂ Aθ 1 1 ∂ Ar ∂ (rAφ ) 1 ∂ (rAθ ) ∂ Ar
∇×A = − r̂ + − θ̂ + − (1.33)ϕ̂
r sin θ ∂θ ∂φ r sin θ ∂ φ ∂r r ∂r ∂θ
∂ 2A
   
2 1 ∂ 2∂A 1 ∂ ∂A 1
∇ A = 2 r + 2 sin θ + 2 2 . (1.34)
r ∂r ∂r r sin θ ∂ θ ∂θ r sin θ ∂ φ 2

Example 1.9: Let ⃗r(x, y, z) = x⃗i + y⃗j + z⃗k be the position vector field on R3 . Then
∥⃗r(x, y, z)∥2 =⃗r ·⃗r = x2 + y2 + z2 is a real-valued function. Find

(a) the gradient of ∥⃗r∥2

(b) the divergence of⃗r
(c) the curl of⃗r
(d) the Laplacian of ∥⃗r∥2

Answer:
(a) ∇∥⃗r∥2 = 2x⃗i + 2y⃗j + 2z⃗k = 2⃗r

∂ ∂ ∂
(b) ∇ ·⃗r = (x) + (y) + (z) = 1 + 1 + 1 = 3
∂x ∂y ∂z
 
 î
 jˆ k̂ 
∂ ∂ ∂ 
(c) ∇ ×⃗r =  = 0î − 0 jˆ + 0k̂ = ⃗0
∂x ∂y ∂ z 
 x y z 

∂2 2 ∂2 2 ∂2 2
(d) ∇∥⃗r∥2 = (x + y2
+ z2
) + (x + y2
+ z2
) + (x + y2 + z2 ) = 2 + 2 + 2 = 6
∂ x2 ∂ y2 ∂ z2
 1.3 Mathematical Symbols 15

1.3 Mathematical Symbols

1.3.1 Kronecker Delta and Levi–Civita Symbols
The Kronecker and the Levi–Civita symbol are represented by δi j and εi jk respectively,
where the indices can take the values x, y, z or 1,2,3 for cartesian and other generalized
coordinates respectively. These functions are defined by
(
1, if i = j
δi j = (1.35)
0, if i ̸= j
and

1,
 if i jk = xyz yzx zxy
εi jk = −1, if i jk = xzy zyx yxz (1.36)

0, otherwise.


These are particularly useful since expressions in cartesian and even other coordinate
systems can be easily written out as well as evaluated using the above definitions, as
discussed next.

1.3.2 Uses of Kronecker Delta and Levi–Civita symbol

Let us define a notation to make writing down equations easier for us:
∂
∂k ≡ (1.37)
∂ xk
Einstein introduced the following convention, which we will now employ. An index which
appears exactly twice in a mathematical expression is implicitly summed over all possible
values for that index. The range of this dummy index must be clear from context and the
index cannot be used elsewhere in the same expression for another purpose. In this book,
the range for an index like i is from 1 to 3, indicating a sum over the cartesian indices x, y,
and z. Thus, a dot product between two vectors A and B is written as
3
A·B = ∑ Ak Bk = Ak Bk (1.38)
k=1

Using this, we can write down divergence and curl respectively,

∇ · A = ∂k Ak (1.39)
[∇ × A]i = εi jk ∂ j Ak (1.40)
Using the following identities, we can now evaluate some vector identities:
êi · ê j = δi j (1.41)
δkk = 3 (1.42)
∂ j rk = δk j (1.43)
A k δk j = Aj (1.44)
δi j εi jk = 0 (1.45)
εi jk εi jk = 6 (1.46)
εi jk εist = δ js δkt − δ jtδ ks (1.47)
 16 Chapter 1. Mathematical Preliminaries

Let us now use the above for evaluating a couple of vector formulae.

[∇ × (A × B)]i = εi jk ∂ j (A × B)k = εi jk εkst ∂ j (As Bt ) (1.48)

= εki j εkst ∂ j (As Bt ) = (δis δ jt − δit δ js )(Bt ∂ j As + As ∂ j Bt ) (1.49)
[∇ × (A × B)]i = Ai ∂ j B j − A j ∂ j Bi + B j ∂ j Ai − Bi ∂ j A j (1.50)
∇ × (A × B) = A(∇ · B) − B(∇ · A) + (B · ∇)A − (A · ∇)B (1.51)

Example 1.10: Prove vector identity ∇ × (∇ × ⃗A) = ∇(∇ · ⃗A) − ∇2⃗A using Levi–Civita
symbol.
Answer: We’ll write the ith Cartesian component of the gradient operator ∇ as ∂i (cf. ∂∂xi ).
Let’s simplify ∇ × (∇ × ⃗A(x)). We start by considering the ith component and then we use
our expression for the cross product (working from the outside in):

∇ × (∇ × ⃗A)i = εi jk ∂ j (∇ × ⃗A)k

Next we replace the remaining cross product, making sure to introduce new dummy
summation variables l and m:

∇ × (∇ × ⃗A)i = εi jk ∂ j εklm ∂l Am = εi jk εklm ∂ j ∂l Am

(The partial derivatives act only on the components of A, so we can pull out the ε’s.) We
rotated the indices in one of the ε’s in the last step so that we can now directly apply our
very useful identity (and simplify)

(∇ × (∇ × A))i = (∂il ∂ jm − ∂im ∂l j )∂ j ∂l Am = ∂m ∂i Am − ∂l ∂l Ai = ∂i (∂m Am ) − (∂l ∂l A)i

Thus we arrive at

∇ × (∇ × ⃗A) = ∇(∇ · ⃗A) − ∇2⃗A

1.4 Dirac Delta Function

The Dirac Delta function can be defined through its effect on functions in the following way.
Suppose δ (x − x0 ) is the Dirac Delta function, then for any continuous and differentiable
function f (x),
ˆ ∞
f (x)δ (x − x0 )dx = f (x0 ) (1.52)
−∞

In other words, we need a function that "chooses" a point and nothing else; i.e. it should
have a large value at one point only, and should be zero everywhere else, such that it is
normalized, i.e.
ˆ ∞
δ (x − x0 )dx = 1. (1.53)
−∞
1.4 Dirac Delta Function 17

Eq.(1.52) is a filtering property and can be used to prove the following identities
1
δ (ax) = δ (x) for a ̸= 0 (1.54)
|a|
ˆ ∞ 
d ′

d 
dx f (x) δ x − x = − f (x) (1.55)
−∞ dx dx ′
ˆ ∞ x=x
′ 1 ′
dkeik(x−x )


δ x−x = (1.56)
2π −∞
1
δ [g(x)] = ∑ ′ δ (x − xn ) where g(xn ) = 0, g′ (xn )(1.57)
̸= 0.
n |g (x n )|

The dirac delta function is particularly useful in dealing with point charges, where the
charge density is huge at a single point and zero everywhere else. Since we are going to
deal with point charges in 3 dimensions, it makes sense to define the dirac delta function
in 3 dimensions as well, in the following way
ˆ (
1, if r ′ ∈ V
d 3 r f (r)δ (r − r ′ ) = (1.58)
V 0, if r ′ ∈
/V

Problems:
1. Under what circumstances are the vectors (⃗A × ⃗B) × C
⃗ and ⃗A × (⃗B × C)
⃗ equal?
⃗
2. If the vector ⃗A has constant magnitude, show that the vectors ⃗A and ddtA are perpen-
⃗
dicular, provided ddtA ̸= 0.
3. If V is a scalar function of x and y, show that the divergence of the vector field
⃗F = aˆz × grad V is always zero.
4. The direction of a vector ⃗A is radially outwards from the origin, and its magnitude is
krn , where

r2 = x2 + y2 + z2

Find the value of n for which div ⃗A = 0.

5. A vector ⃗A is given in cylindrical coordinates as

⃗A = aˆr 2 cos φ + aˆφ r

Evaluate the line integral of A around the contour in the z = 0 plane bounded by +
x-axes and + y-axes and the arc of the circle of radius 1 unit. Check the answer by
performing the appropriate surface integral of curl ⃗A.
6. A point P moves so that its position vector r relative to another point O satisfies the
equation

d⃗r
⃗ ×⃗r
=ω
dt
where ω⃗ is a constant vector. Prove that P describes a circle with constant velocity.
7. Given ⃗A = 4ŷ + 10ẑ and ⃗B = 2x̂ + 3ŷ, find the projection of ⃗A on ⃗B.
18 Chapter 1. Mathematical Preliminaries
√
8. Given ⃗A = (10/ 2)(x̂ + ẑ) and ⃗B = 3(ŷ + ẑ), find the projection of ⃗B on ⃗A as a vector
in the direction of ⃗A.
9. Find the relationship which the cartesian components ⃗A and ⃗B must satisfy if the
vector fields are everywhere parallel.
10. Express the unit vector directed toward the origin from an arbitrary point on the line
described by x=0, y=3.
11. Given the plane 4x + 3y + 2z = 12, find the unit vector normal to the surface in the
direction away from the origin.
12. Express the unit vector directed toward the point (0, 0, h) from an arbitrary point in
the plane z = −2.
13. Find the unit vector directed from (2, −5, −2) toward (14, −5, 3).
14. Show that the absolute value of ⃗A · ⃗B × C ⃗ is the volume of the parallelopiped with
edges ⃗A, ⃗B and C.
⃗
15. Given ⃗A = 2x̂ − ẑ, ⃗B = 3x̂ + ŷ, and C ⃗ = −2x̂ + 6ŷ − 4ẑ, show that C
⃗ is ⊥ to both ⃗A
⃗
and B.
16. Convert the following rectangular Cartesian coordinate (1, 3, 5) into an equivalent
(a) cylindrical and (b) spherical coordinates. ´
17. Let ⃗F = ⟨xy, −xy⟩ and 0 let D be given by 0 ≤ x ≤ 1, 0 ≤ y ≤ 1. Compute ∂ D ⃗F · d⃗r
´
and ∂ D ⃗F · n̂ ds.
18. If ∇ · ⃗F = 0, ⃗F is said to be incompressible. Show that any vector field of the form
⃗F(x, y, z) = ⟨ f (y, z), g(x, z), h(x, y)⟩ is in-compressible. Give a non-trivial example.
19. Compute ∇ · ⃗F and ∇ × ⃗F
⃗F = (2y − cos (x)) x̂ − z2 e3x ŷ + x2 − 7z ẑ

20. Compute ∇ · ⃗F and ∇ × ⃗F

⃗F = − (4y − 1) x̂ + xy2 ŷ + (x − 3y) ẑ

21. Compute ∇ · ⃗F and ∇ × ⃗F

2
⃗F = z2 (y − x) x̂ + 4y ŷ + x2 − 3z ẑ


z3
22. The Cauchy-Schwarz inequality states that, for any two vectors u and v in Rn :

(u · v)2 ≤ (u · u)(v · v)

with equality holding if and only if u = λ v for some λ ∈ R. Prove the Cauchy-
Schwarz inequality. [Hint: Use angles.]
23. Show that the equation a·r = a2 defines a two dimensional plane in three dimensional
space, where a is the minimal length vector from the origin to the plane. [Hint:
A plane is the translate of the linear span of two vectors. The Cauchy-Schwarz
inequality may come in handy.]
24. The volume of a tetrahedron is V = bh/3, where b is the area of a base and h is the
height (distance from base to apex). Consider a tetrahedron with one vertex at the
origin and the other three vertices at positions ⃗A, ⃗B and C.
⃗ Show that we can write

1
V = ⃗A · (C
⃗ × C)
⃗
6
1.4 Dirac Delta Function 19

This demonstrates that the volume of such a tetrahedron is one sixth of the volume
of the parallelepiped defined by the vectors ⃗A, ⃗B and C.
⃗
25. Using index notation, prove Lagrange’s identity:

(⃗A × ⃗B) · (C
⃗ × ⃗D) = (⃗A · C)(
⃗ ⃗B · ⃗D) − (⃗A · ⃗D)(⃗B · C)
⃗

26. Either directly or using index methods (Levi-Cevita symbols), show that, for any
scalar field φ and vector field ⃗A,
(a) ∇ · (φ ⃗A) = ∇φ · ⃗A + φ ∇ · ⃗A,
(b) ∇ × (φ ⃗A) = ∇φ × ⃗A + φ ∇ × ⃗A, and
(c) ∇2 (φ ψ) = (∇2 φ )ψ + 2∇φ · ∇ψ + φ ∇2 ψ
27. Show that, for r ̸= 0,
(a) ∇ · r̂ = 2r , and
(b) ∇ × r̂ = 0
28. A function φ satisfying ∇2 φ = 0 is called harmonic
(a) Using Cartesian coordinates, show that φ = 1/r is harmonic, where r = (x2 +
y2 + z2 )1 /2 ̸= 0.
(b) Let α = (α1 , α2 , α3 ) be a vector of non-negative integers, and define |α| =
α1 + α2 + α3 . Let ∂α be the differential operator ∂xα1 ∂yα2 ∂zα3 . Prove that any
function of the form φ = r2|α| +1 ∂α (1/r) is harmonic. [Hint: Use vector calculus
identities to expand out ∇2 (rn f ) where f := ∂α (1/r) and use Euler’s theorem and
the fact that mixed partials commute.]
 2. Electrostatics

2.1 Conducting Sphere in a Uniform Electric Field

Consider a perfectly conducting sphere placed in a uniform electric field (magnitude E0 ),
pointing in the +z direction. Find
1. Electric field in the region
2. Potential in the region
3. Surface charge density on the sphere and
4. Induced dipole moment of the sphere
−→ FIRST: Write down ALL boundary values:
1. At r = R, V =constant
=⇒ at r = R, W.L.O.G., we can take V =0

2. Far away from the sphere, i.e. for r −→ ∞, E −→ E0 ẑ

E = E0 ẑ
∂V
=⇒ − = E0
∂z
=⇒ V = E0 z = −E0 r cos θ

−→ Now, let us move to symmetry in the problem...

−→ Extremely important to utilize all symmetries of the problem...
−→ Here, clearly, there is a symmetry about the angle φ
−→ Now, choose an appropriate co–ordinate system . . .
−→ Of course, a different co–ordinate system will ALSO work; it will just require more
ALGEBRA, that is all . . .
22 Chapter 2. Electrostatics

−→ Now, write down the relevant Electrostatics equation:

∇2V = 0 (2.1)

−→ Since there is a symmetry around φ , so the Laplacian of V can be written as

   
2 1 ∂ 2 ∂V 1 ∂ ∂V
∇V= 2 r + 2 2 sin θ (2.2)
r ∂r ∂r r sin θ ∂ θ ∂θ
−→ Now, as before, we write

V (r, θ ) = Z(r)P(θ ), (2.3)

and get, after some algebra,

   
1 d 2 dZ 1 d dP
r + sin θ =0 (2.4)
Z dr dr P sin θ dθ dθ

Now, substituting x = cos θ ,

d d
= −
sin θ dθ dx
d p d
=⇒ = − 1 − x2
dθ dx
−→ so that, on equating the first expression on the LHS to ℓ(ℓ + 1) and the second one
to −ℓ(ℓ + 1) (where ℓ MUST be an integer), we get
 
1 d 2 dZ
r = ℓ(ℓ + 1) (2.5)
Z dr dr
d2P dP
(1 − x2 ) 2 − 2xy + ℓ(ℓ + 1) = 0 (2.6)
dx dx
2.1 Conducting Sphere in a Uniform Electric Field 23

−→ The second equation on the previous slide is known as the Legendre Equation
−→ The solution to the first one is
B
Z(r) = Arℓ + (2.7)
rℓ+1
where the constants A and B are to be determined from Boundary conditions.

The solution to the second equation is

P(θ ) = Pℓ (cos θ ) (2.8)

So, the net solution then is

∞  
ℓ Bℓ
V (r, θ ) = ∑ Aℓ r + ℓ+1 Pℓ (cos θ ) (2.9)
ℓ=0 r

−→ Note that we MUST sum over all possible values of ℓ

−→ We have, in principle, obtained a solution and we can verify that it satisfies the Laplace
equation, therefore, by the Uniqueness Theorem, it MUST be the ONLY solution.
−→ But we still need to apply Boundary conditions to determine all the Aℓ s and the
Bℓ s
−→ To do that, note the first Boundary condition:
At r = R, we must have V = 0
−→ Thus, only the Z(r) part of the solution must go to zero at r = R, so that
Bℓ
Aℓ Rℓ + = 0
Rℓ+1
=⇒ Bℓ = −Aℓ R2ℓ+1 (2.10)

so that the solution reduces to

R2ℓ+1
∞  
ℓ
V (r, θ ) = ∑ Aℓ r − ℓ+1 Pℓ (cos θ ) (2.11)
ℓ=0 r

−→ Let’s now apply the second Boundary condition that

at r = ∞, V = −E0 cos θ
−→ Obviously, this concerns ONLY the P(θ ) part of the solution, so let’s look at how to
get cos θ from Pℓ (cos θ )

−→ Turns out, this is SIMPLE:

P0 (cos θ ) = 1
P1 (cos θ ) = cos θ
1
3 cos2 θ − 1


P2 (cos θ ) =
2
.....
i.e. there is ONE AND ONLY ONE WAY TO GET P(θ ) = cos θ and that is when ℓ = 1
24 Chapter 2. Electrostatics

−→ We DON’T need to sum over all ℓs, so that

R3
 
V (r, θ ) = A1 r − 2 cos θ (2.12)
r

−→ Let us now take the limit r −→ ∞:

A1 r cos θ = −E0 r cos θ =⇒ A1 = −E0 (2.13)

FINALLY FINAL SOLUTION:

R3
   3 
R
V (r, θ ) = −E0 r − 2 cos θ = E0 − 1 r cos θ (2.14)
r r3

−→ BUT . . . we are not yet done . . .

−→ Need to find Electric Field as well . . .
−→ It turns out that for this co–ordinate system,
R3
 
∂V
Er = − = E0 1 + 2 3 cos θ (2.15)
∂r r
R3
 
1 ∂V
Eθ = − = E0 1 − 3 sin θ (2.16)
r ∂θ r
Also, the surface charge density

σ (θ ) = ε0 Er |r=R = 3ε0 E0 cos θ (2.17)

 2.2 Green’s Function Method 25

2.2 Green’s Function Method

Let us write the generalized Poisson equation

∇2 ψ(r) = −4π f (r) (2.18)

The Green’s function method works as follows:

For the Green’s function

∇2 G(r, r ′ ) = −4πδ (r − r ′ ), (2.19)

the solution to the above Poisson equation is

ˆ
ψ(r) = ψ0 (r) + G(r, r ′ ) f (r ′ )d 3 r′ (2.20)

Here,
−→ ∇2 ψ0 (r) = 0 −→ The source term
−→ So how does it work?
Consider eq.(2.19); multiply both sides with f (r ′ ) and integrate over r ′ :
ˆ ˆ
∇ G(r, r ) f (r ) d r = −4π f (r ′ )δ (r − r ′ )d 3 r′
′ ′
2
  3 ′
(2.21)

where we can take f (r ′ ) inside ∇2 since the operator is only a function of r and not r ′ .
We then get
ˆ 
2 ′ ′ 3 ′
∇ G(r, r ) f (r )d r = −4π f (r) = ∇2 ψ(r) (2.22)

Therefore, in the previous equation, ψ(r) must be equal to the expression within the square
brackets, PLUS another term, the Laplacian of which is zero...

...so that
ˆ
ψ(r) = ψ0 (r) + G(r, r ′ ) f (r ′ )d 3 r′ (2.23)

So now, we must find G(r, r ′ ) through the equation that defines it:

∇2 G(r, r ′ ) = −4πδ (r − r ′ ) (2.24)

At first glance, it seems like this is going to be equally difficult as the Poisson equation
solution itself without this method!
−→ However, the reason it is simplified is that we can write the Green’s function in terms
of its fourier transform:
ˆ
′ 1 ′

3
G(r, r ) = G̃(k) exp ik · (r − r ) d k (2.25)
(2π)3
−→ By writing the Green’s function in terms of its fourier transform, we can change
the equation we wish to solve from a differential equation to an algebraic equation in the
26 Chapter 2. Electrostatics

following way
First, substitute the expression for δ (r − r ′ ):
ˆ
′ 1 ′

3
δ (r − r ) = exp ik · (r − r ) d k (2.26)
(2π)3
and then substitute for G(r, r ′ ) in eq.(2.24) to get
ˆ
∇ G̃(k) exp ik · (r − r ) d k = −4π exp ik · (r − r ′ ) d 3 k
2 ′

3 

(2.27)

Now, remember that the operator ∇2 is only a function of r and not r ′ , so that it ONLY
operates on the exponential inside the integral on the LHS of eq.(2.27)
2
−→ Also, when it acts on the exponential inside the integral, it is effectively ∂∂r2
−→ So, performing the differentiation twice on the exponential inside the integral on the
LHS, we get
ˆ ˆ
k G̃(k) exp ik · (r − r ) d k = 4π exp ik · (r − r ′ ) d 3 k
2 ′

3

(2.28)

Therefore, it must be that

k2 G̃(k) = 4π (2.29)
−→ Now, all we need to do is substitute the value of G̃(k) in eq.(2.25) to get
ˆ ik·(r−r′ )
′ 1 e
G(r, r ) = 3
4π d3k (2.30)
(2π) k2
To do this, make the following substitutions:
kx = k sin θ cos φ (2.31)
ky = k sin θ sin φ (2.32)
kz = k cos θ (2.33)
k · (r − r ′ ) = kR cos θ (2.34)
d3k = k2 sin θ dkdθ dφ (2.35)
where R = |r − r ′ |
After all the substitutions, we get
ˆ ˆ ˆ ikR cos θ
′ 1 e
G(r, r ) = 2 2
k2 sin θ dkdθ dφ (2.36)
2π k
which yields, after some algebra and one more substitution:
1 1
G(r, r ′ ) = ≡ (2.37)
R |r − r ′ |
Therefore, the solution to the Poisson equation is
ˆ
f (r ′ ) 3 ′
ψ(r) = ψ0 (r) + d r (2.38)
|r − r ′ |
ρ(r)
If 4π f (r) = ε0 , then
ˆ
1 ρ(r ′ ) 3 ′
ψ(r) = ψ0 (r) + d r (2.39)
4πε0 |r − r ′ |
which is exactly what we expect intuitively.
3. Conductors, Dielectrics and Capacitan
 4. Magnetostatics and Magnetic Forces

We have seen in §2 on Electrostatics that static charges exert a force on each other, which
is proportional to the inverse squared distance between them, and to the multiple of the two
charges. While it is reasonable to expect that such a force would also exist between moving
charges, applying Coulomb’s Law to moving charges is not possible. Therefore, for charges
moving with a constant speed; in other words, for steady currents, we need to describe
a different force, which is referred to as the magnetic force. Thus, while electrostatics
deals with static electric charges, magnetostatics deals with stationary electric currents, i.e.,
electric charges moving with constant speeds, and the interaction between these currents.
The study of magnetic forces or magnetism has historically involved the use of what are
called magnetic materials. We shall deal with these in the next chapter; for the discussion
in this chapter, we will concentrate on wires carrying steady currents. As we shall see
later, this simplifies our treatment and understanding of magnetic forces, i.e. magnetism
considerably.

4.1 Force between current carrying wires: Biot–Savart Law

Experiments by Coulomb showed

4.2 Divergence and Curl of the Magnetic Field

µ0 (r − r ′ )
dB = Idl × (4.1)
4π |r − r ′ |3
but since

Idl = jdV ′ , (4.2)

30 Chapter 4. Magnetostatics and Magnetic Forces

we get
µ0 (r − r ′ ) ′
dB = j(r ′ ) × dV (4.3)
4π |r − r ′ |3
−→ Now,

∇ × ( f A) = f (∇ × A) + A × ∇ f (4.4)

so, using this in the previous equation, with

1
f =
|r − r ′ |
A = j(r ′ )
we get
ˆ
µ0 j(r ′ )
B = ∇× dV ′ (4.5)
4π |r − r ′ |
Define this as a Vector PotentialA

since the other term reduces to zero (because the ∇ operator is a function of r only whereas
j is a function of r ′ only)

So that we get, finally,

ˆ
µ0 j(r ′ )
B = ∇× dV ′ (4.6)
4π V ′ |r − r ′ |
Also,

∇ · (A × B) = B · (∇ × A) − A · (∇ × B) (4.7)

so that
ˆ " #
µ0 (r − r ′ )
∇·B = ∇ · j(r ′ ) × dV ′ (4.8)
4π V′ |r − r ′ |3
And so
ˆ " #
µ0 (r − r ′ )
∇·B = ∇ · j(r ′ ) × dV ′
4π V′ |r − r ′ |3
ˆ  
µ0 ′ 1
= − j(r ) · ∇ × ∇
4π V′ |r − r ′ |
≡ 0 (4.9)
−→ “No magnetic monopoles” is built–in to Biot–Savart Law!
This also makes sense if we define
B ≡ ∇ × A where
ˆ
µ0 j(r ′ )
A = dV ′ (4.10)
4π |r − r ′ |
4.2 Divergence and Curl of the Magnetic Field 31

Exercise: Show that

∇ × B = µ0 j(r) (4.11)

using vector identities.

The above is equivalent to
˛
B · dl = µ0 I (4.12)
L

Now, force on moving charge due to magnetic field alone:

F = qv × B (4.13)

so that the net EM force is

F = q (E + v × B) (4.14)
5. Magnetic Materials and Inductance
 6. Time Varying Fields and Maxwell’s Equa

6.1 Issue with Ampere’s Law

There is clearly an issue with Ampere’s Law; on taking the divergence of both sides:

∇ · ∇ × B ≡ 0 = µ0 ∇ · j (6.1)

but Ampere’s Law MUST hold for all current densities, not only divergenceless ones!
Turns out, conservation of charge saves us here, in the following way
For any volume V , the sum total of charges moving in and out must be exactly the same:
˛ ˆ 
∂ ∂ρ
− j · dS = ρdV =⇒ ∇ · j + =0 (6.2)
∂t ∂t

So we can now substitute for the divergence of the current density thus:
 
∂ ∂E
∇ · j = −ε0 [∇ · E] =⇒ ∇ · j + ε0 (6.3)
∂t ∂t

=⇒ we must replace j with j + ε0 ∂∂tE in Ampere’s law to get

∂E
∇ × B = µ0 j + µ0 ε0 (6.4)
∂t

6.2 Statics −→ Dynamics

Faraday’s Law of Induction (paraphrased) −→ “A time–varying magnetic field induces an
electric field”
More concrete: Induced potential across circuit is equal in magnitude to the rate of change
 Chapter 6. Time Varying Fields and Maxwell’s Equations: Electrodynamics
36

of magnetic flux, i.e.

˛ ˆ
∂
E · dl = − B · dS (6.5)
∂t
or, equivalently, using the definition of the curl:
 
∂B ∂A
∇×E = − = ∇× − (6.6)
∂t ∂t

This CLEARLY suggests that we can write the electric field consists of 2 parts and can
therefore be written as
∂A
E = −∇ϕ − (6.7)
∂t
and the magnetic field is

B = ∇×A (6.8)

Let us now use the four Maxwell equations in their dynamic form
ρ
∇·E = (6.9)
ε0
∂B
∇×E = − (6.10)
∂t
∇·B = 0 (6.11)
∂E
∇ × B = µ0 j + µ0 ε0 (6.12)
∂t
and substitute in the above to get, after some algebra,

1 ∂ 2ϕ
 
2 ρ ∂ 1 ∂ϕ
−∇ ϕ = + ∇·A+ 2 (6.13)
c2 ∂t 2 ε0 ∂t c ∂t
1 ∂ 2A
 
2 1 ∂ϕ
− ∇ A = µ0 j − ∇ ∇ · A + 2 (6.14)
c2 ∂t 2 c ∂t
The above equations are the Maxwell equations in terms of the potentials. Note that this is
a set of four equations, since A is the vector potential.
Notice that the terms in the square brackets in both the equations are EXACTLY the
same. This is not a coincidence. These represent what is known as a Gauge Freedom, i.e.
the freedom to choose the divergence of the vector potential and the time rate of change of
the scalar potential.

6.3 Non–Uniqueness of Scalar & Vector Potential

−→ Remember how the vector potential is defined

B = ∇×A (6.15)
6.3 Non–Uniqueness of Scalar & Vector Potential 37

due to which we can only define A up to the gradient of a scalar function, say χ, so that
the transformation A → A′ = A + ∇χ produces no difference in the magnetic field B.

However, the electric field does change:

∂ A ∂ ∇χ
E ′ = −∇ϕ − − (6.16)
∂t ∂t
so that
 
′ ∂χ ∂A
E = −∇ ϕ + − (6.17)
∂t ∂t
=⇒ In order to keep the electric field unchanged as well, we need to have the following
replacement:
∂χ
ϕ → ϕ′ = ϕ − (6.18)
∂t
−→ Then, E will remain unchanged as well.
=⇒ We can actually CHOOSE the function χ(r) in the alternate vector potential
A′ (r) = A(r) + ∇χ(r) (6.19)
In practice, we always put a CONSTRAINT on the vector potential A(r), which amounts
to choosing χ(r):
∇ · A(r) = 0, or, alternately, ∇ · A′ (r) = 0 =⇒ ∇2 χ(r) = −∇ · A(r) (6.20)
=⇒ at worst, we need to solve the Poisson equation for χ(r) with ∇ · A(r) as the
source term . . .

−→ So how does this help us?

Let us consider Ampere’s Law
∇ × B = ∇ × (∇ × A) = µ0 j (6.21)
which is the magnetostatic equivalent of the Poisson equation
−→ This is not easy to analyze; if ji is the ithe component of the current density vector
j(r), the above equation implies
∇i (∇ · A) = µ0 ji (6.22)
so that all the three Cartesian components of A(r) are non–trivially combined. −→ The
constraint ∇ · A(r) = 0 provides an easy way out; with this constraint, we have
∇2 A(r) = −µ0 j(r) (6.23)
Now, we can immediately use the Green’s function method for each component of A(r)
to obtain
ˆ ′ ′
µ0 j (r ) 3 ′
A(r) = d r (6.24)
4π |r − r ′ |
This CONSTRAINT is called a Gauge Condition or just Gauge.
7. Electromagnetic Waves
 A. The Frobenius Method: Special Function

The differential equation

d2y dy
1 − x2 2
− 2x + n(n + 1)y = 0 (A.1)
dx dx
is known as Legendre’s equation1 . The parameter n is a real number, and in most
applications, only integral values are needed.
To obtain a solution, assume
∞
y= ∑ ak xm+k (A.2)
k=0

which gives, on substitution into eq.(A.1)

1 − x2 ∑(m+k)(m+k −1)ak xm+k−2 −2x ∑(m+k)ak xm+k−1 +n(n+1) ∑ ak xm+k = 0,

(A.3)

or

∑(m + k)(m + k − 1)ak xm+k−2 − ∑{(m + k)(m + k + 1) − n(n + 1)}ak xm+k = 0 (A.4)

Now, the ONLY way that the entire LHS can equal zero is if each power of x equals zero.
Let us look at the coefficient of the lowest power of x, i.e. coefficient of xm−2 , which must
be = 0

m(m − 1)a0 = 0 i.e. m = 0 or m = 1. (A.5)

1 Named
after the French mathematician Adrien Marie Legendre, who made important contributions to
Number Theory and Elliptic Integrals.
42 Chapter A. The Frobenius Method: Special Functions

Now let us examine the next higher power of x, whose coefficients must also be = 0
m(m + 1)a1 = 0 =⇒ a1 = 0; (A.6)
and, combined with the next such relations, we will get a1 = a3 = a5 . . . = 0. Thus,
alternate terms are absent in eq.(A.2).
Now consider eq.(A.4). In order to make the power of x m + k, we can make the
substitution k → k + 2 WLOG (since all ak s are assumed to be zero). This makes the
equation
∑(m+k +2)(m+k +1)ak+2xm+k − ∑{(m+k)(m+k +1)−n(n+1)}ak xm+k = 0 (A.7)
such that we may write
(m + k + 2)(m + k + 1)ak+2 − {(m + k)(m + k + 1) − n(n + 1)}ak = 0
(m + k − n)(m + k + n + 1)
=⇒ ak+2 = ak
(m + k + 1)(m + k + 2)
For m = 0, this becomes
(n − k)(n + k + 1)
ak+2 = ak
(k + 1)(k + 2)
Therefore,
n(n + 1)
a2 = a0 ,
2!
(n − 2)(n + 3) n(n + 1)(n − 2)(n + 3)
a4 = a2 = a0 (A.8)
3·4 4!
etc. The coefficients for the case m = 1 can be similarly calculated.
Hence, the two solutions are
 
n(n + 1) 2 n(n + 1)(n − 2)(n + 3) 4
y1 = a0 1 − x + x +...
2! 4!
 
(n − 1)(n + 2) 3 (n − 1)(n + 2)(n − 3)(n + 4) 5
y2 = a0 x − x + x +... (A.9)
3! 5!
These polynomials, with the value of a0 normalized such that their value is 1 at x = 1, are
known as Legendre Polynomials are are denoted by Pn (x).
Note the following properties of Legendre Polynomials:
1. These series, when non–terminating, are convergent for |x| < 1
2. When n is an integer, one or the other of the two series in eqs.(A.9) terminates
3. Some of the first few Legendre Polynomials are
P0 (x) = 1
P1 (x) = x
1
3x2 − 1


P2 (x) =
2
1
5x3 − 3x


P3 (x) =
2
1
35x4 − 30x2 + 3


P4 (x) =
8
1 5 
P5 (x) = 63x − 70x3 + 15x
8
... (A.10)
 Bibliography

Books
Articles
 Index

C
Conducting Sphere in Uniform Electric Field
21