A FIRST GLIMPSE ON

SPECIAL RELATIVITY
Prasun Dutta
Copyright c 2018 Prasun Dutta

D R . P RASUN D UTTA
Assistant Professor, Department of Physics.
Indian Institute of Techology (BHU), Varanasi

Varanasi, Oct 2018

 Table of Contents

1 Simultaneity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.1 Postulates of Special Relativity 5
1.2 Defining Reference Frames 5
1.3 Space-Time Diagram 7
1.3.1 Defining the axes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.3.2 Events, measurements and world lines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

1.4 Thought experiments 10

1.4.1 A Thought Experiment in a Train Cart . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

2 Lorentz Transformations: Space-Time . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

2.1 Lorentz Transformation Relations 13
2.2 Concept of Interval 14
2.2.1 Light Like Intervals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.2.2 Time Like Intervals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.2.3 Space Like Intervals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

3 Lorentz Transformation: Consequences . . . . . . . . . . . . . . . . . . . . . . . . . 18

3.1 Time Dilation 19
3.2 Length Contraction 21
3.3 Addition of velocities 23
4

4 Dynamics of a free particle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26

4.1 Lagrangian of a free particle 26
4.2 Relativistic Momentum 27
4.3 Relativistic Energy 28
4.4 Relativistic Momentum-energy relation 28
4.5 Final Notes 29
 1. Simultaneity

1.1 Postulates of Special Relativity

We already know the postulates of the special relativity. Let us start from there.
1. Physical Laws are same in all inertial reference frames.
2. Speed of light is constant in free space and does not depend on the speed of the observer or
the source.
We know that if we use the Galilean Transformation relations to calculate how the speed of light
changes in different inertial reference frames, the speed of light appear to be different in different
reference frames. Hence, it is clear that the Galilean Transformation will not hold true with the
postulates of the special relativity.
When we wrote the Galilean transformation relations we assumed that the time interval as measured
by different observers do not change between two inertial reference frames. Let us examine if we can
carry forward the notion of invariant (means constant) time with the postulates of special relativity.

1.2 Defining Reference Frames

Let us first try to understand what we mean by a reference frame. A usual misconception is to
consider only one person (we call observer) to do all the measurements from a reference frame in
whatever way he likes. Though such single observer measurement rules can be figured out, it is not
very straight forward. We shall rather choose a different way of defining reference frames.
We consider an infinite number of observers, who have volunteered to form one reference system. At
the beginning of an experiment, we call them to meet in a cafe and discuss the protocols or rules of
the measurement. When they arrive, we give each of them a piece of paper. On the top of the paper,
there are three numbers. These three number combination is different for all the different observers.
Only one of the observers gets a special paper with no number on it. This special observer has a free
will, he can go to any position or point in space and stand their. We call that position as origin of our
reference system. The three numbers all the other observers have essentially points to a location
6 Chapter 1. Simultaneity

Figure 1.1: Defining a reference frame with many observers. All the observers are given a piece
of paper and a clock. The clocks are all synchronized. There is a special observer with the paper
blank. Everybody else have a number written on their paper. They take their position according to
the number written on the paper. This set of observers constitutes a reference frame S.

in space which is at a certain distance away from our observer at the origin. Each of the observers
are also given very accurate clocks. All the observers match the time of their clock to our special
observer. They set the time to zero and let the clock run. This process is called synchronization of
the clocks. Now, they are told to do the following things:
• All the observers will go and stand at the positions marked in their piece of paper.
• They must not be moving even by a tiny amount after they reach their position.
• If something happen at their location, they will look at their clock and note down the time in
the piece of paper they have.
• These observers would be asked to come back to the cafe again after the measurement session
is over. They will compare their papers in the cafe and get the results.
The collection of these observers makes one inertial reference frame. Let us give this reference
frame a name for convenience. We call it the reference frame S. If we have to make another inertial
reference frame, say S0 we would need another infinite set of observers, we will do the same drill of
giving measurement papers to each observer, choose a special observer, give them clocks etc etc. The
new observers would also follow the above set of rules. The only difference is that all the observers
that makes the reference frame S0 are moving with a velocity with respect to all the observers from
the frame S.
That is how they will constitute a different reference frame S0 . You can constitute as many reference
frame you want this way.
Now that we understand what we mean by a reference frame and how the observations are performed,
let us try to have a pictorial representation of these.
 1.3 Space-Time Diagram 7

1.3 Space-Time Diagram

It helps us in many ways to use a diagrammatic approach to see what is happening through the boring
equations. Since we may now also have to deal with time explicitly when we consider transformation
between different coordinates, we have a total of four variables to look at. That is the three space and
a time variable. The transformations between different reference frames we are interested in here are
induced by relative velocity between them. In general this velocity can have x, y or z components.

1.3.1 Defining the axes

Unfortunately, we can draw only two dimensional diagrams in a piece of paper. Hence, let us
simplify the problem here. We sacrifice two spacial dimensions and assume that the relative velocity
between the reference frames is only along the x direction. 1
Now, with only one space and one time dimension, we can have a pictorial representation of what is
happening in a piece of paper. We use the horizontal axis to draw the space dimension, we call it x.
The vertical axis we use to draw time. However, to give the vertical axis the same dimension as the
horizontal axis (that is of space or length dimension), we use ct as our vertical axis.

1.3.2 Events, measurements and world lines

let us consider that we are in charge of the observers in the reference frame S. If anything is seen to
happen in a given position x1 at a given time ct1 in S, we call it an event. We may label the event as
E = [ct1 , x1 ] for convenience 2 . Clearly, this marks a point in our space-time diagram.
Let us now ask how is this event recorded in reference frame S. Try to remind yourself of the
measuring protocols we had. There is one observer at the position x1 . When at the time ct1 the event
happens at his place, he notes down the time in his piece of paper.
Let us now consider a little more complicated set of measurements. Let us ask the observers in
reference frame S to measure the relative velocity of one observer in the reference frame S0 . Say
the observer is called Flash. Flash can only move at a constant speed with respect to the observes
in the reference frame S. Jay and Jaya are two observers in the frame S who records two events
EA = [ctA , xA ] and EB = [ctB , xB ]. These events are the events of Flash going past Jay and Jaya.
As the measurement protocol suggests, Jay and Jaya will meet in a cafe. They will compare their
readings from their piece of observation papers. They already had their location xA and xB written
on top of their papers. During the measurement they have noted down ctA and ctB . They can now
calculate the velocity of Flash by doing
xB xA
v= . (1.1)
tB tA
1 We can do this assumption without loss of any generality. A basic assumption in Physics is that the space does not

have inherent directional property, i.e, the physical laws does not change if you change the direction of the instrument you
are using to measure. The set of observers we defined, does not have any equipment to know which direction they are
looking at. They measure only things that are happening next to them. They are called local observers. Hence a given
reference frame inherently can not have any special direction called x or y or z. They can always define the direction x
along the direction of motion of the object they are measuring the velocity of. Well, if the object changes its velocity and
move in different direction at different times, this simplification will not work. But let us first understand simple things.
2 Observe that here we have written the vertical variable first in the ordered pair. This you can take as a definition and is

written this way just as a convention.

8 Chapter 1. Simultaneity

Figure 1.2: In the left figure an event is shown with a small circle. The observer who measure the
event is at the position where the event happens. In right Jay and Jaya measures the events when
Flash pass by them. The world line of flash is shown with a dark dashed line.

Let us now try to draw the trajectory of Flash in the space-time diagram. Remember Flash is moving
with a constant velocity along the x direction. If he was at x = x0 at t = 0, at a later time t, his
position x(t) would be
v v
x(t) = x0 + ct = x0 + b ct, b= (1.2)
c c
Clearly, all the events in the trajectory of Flash would lie in a straight line in our space-time diagram
with a slope of cv . 3
Trajectories of any particles or anything drawn in the space time diagrams are called the world lines.
Let us take a few examples of world lines to understand it better.
• A particle at rest at x = x0 : The velocity of the particle is zero. We can consider that this
particle is moving with a velocity zero. Hence the world line of this particle will be a straight
line with the slope of the line infinity. That is a vertical line passing through x = x0 (see figure
1.3-a).
• A particle moving with a constant velocity v along x axis: Just like the above, only the slope
will be cv (see figure 1.3-b).
• Tip of a pendulum of time period T , mean position x = x0 and maximum amplitude x1 : Here,
the tip of the pendulum does not move with a constant velocity. At the position of extreme
amplitude at x = x0 + x1 , the pendulum has zero velocity and the velocity decreases as it comes
towards its mean position x0 . We still do not know how to draw such world lines. In this case,
3 Remember to find the slope of the straight line we mentioned here, you have to invert the above equation and write ct
in terms of x and hence the slope is cv .
1.3 Space-Time Diagram 9

Figure 1.3: World-lines for the examples given in the text. We draw the world line of light with blue
lines.
 10 Chapter 1. Simultaneity

we can consider the world line of the pendulum in very small intervals of time. If the time
interval is small enough, we may say that it is moving in a straight line during that interval.
The slope of the line will be the speed of light c divided by it’s instantaneous velocity. Now
we can draw such world line (see figure 1.3-c).
• A light ray started by the observer at the origin towards positive x axis: Light rays will have
similar trajectory as that of a particle. But there is a special thing for the light rays, they have
the velocity same as c. Hence, slope of their world lines are always ± cc = ±1 (see figure
1.3-d).
Problem 1.1 Draw the world line of a particle thrown vertically upwards from the earth’s surface
with a speed lesser than the escape velocity.

1.4 Thought experiments

Galileo found the transformation relations between two inertial reference frames S and S0 by doing
real experiments. The relative velocities he could use between the reference frames were a few
meters per second. A modern day Jet Airliner travels at a speed of about 700 km hr 1 , which is
about 200 m s 1 (see figure 1.2). All these are examples of velocities much less that of the speed of
light c = 299792458 m s 1 . We will see in the next chapter, that for all such speed, much smaller
compared to the speed of light, the Galilean laws of transformation holds quite good. However, if
there are motion near to the speed of light, then they need to be modified.
To investigate what happens for motion near to the speed of light we need to do experiments with
velocities of particles near to the speed of light. Such experiments are not easy to perform in a regular
laboratory. Einstein invented thought experiments to analyze such motions. These experiments are
not performed in a laboratory, rather they happen in our brain. These are like the numerical problems:
What distance would a cannon ball travel if it is thrown with a velocity 50 m s 1 at an angle 30
with the horizontal? To solve this problem, you do not actually throw the cannon ball and measure
how much distance it travel. You assume Newton’s laws of motions work and use that to calculate
the distance. In the thought experiments Einstein considered that the two postulates of relativity
holds and tried to see what would be the result of these experiments.

1.4.1 A Thought Experiment in a Train Cart

Consider a train cart. The length L of the train cart is so large that it takes several seconds for the
light to reach from one end to the other! The observers standing inside the train cart defines our
reference frame S. They are all at rest with the train. For this particular experiment we are going to
conduct, we need only three observers. One of them is standing at the middle of the train at x = x0 ,
we call her P. One observer is at the left end of the train, that is at x = x0 L. We call him A. The
third observer is at the right end of the train at x = x0 + L, he is called B. As time passes their world
lines are vertical lines in space-time diagram separated by length L for AP and PB (see the black
dashed line in figure 1.4).
We have given a double sided flashlight (torch) to P. Once it is switched on, light rays come out from
two bulbs at two sides of it and goes towards +x and x directions. Let us assume that she switches
her torch on at a time t = 0 in her clock. We call this event EP = [0, x0 ].
How would be the world line of the light rays going in +ve and ve x direction? As we have already
seen they will have slope of 45 s and 135 respectively. Let us draw these lines in the space time
1.4 Thought experiments 11

Figure 1.4: Dark dashed lines are the world line of the observers A, P and B. The blue continuous
line are the world line of light rays. The events EA and EB happen at the intersection of the world
line of A and B and the light rays. These events are simultaneous in the reference frame S but not in
the reference frame S0 .

diagram (blue lines in figure 1.4). The observers A and B would record the events when the light ray
reaches them. These events are the intersection of the world line of these observers with the world
line of the corresponding light rays. Let us call these events EA and EB .
The question is what would be the time that the observers A and B notes down? We can easily
calculate that using the second postulates. The light in both case has to travel a distance L. Hence,
in both case the time that A and B would note down will be Lc . The events EA = [L, x0 L] and
EB = [L, x0 + L]. These two events happen exactly at the same time, we say the events EA and EB are
simultaneous in reference frame S.
Now let us consult the observers in a different reference frame about the same two events. The
observers in the new reference frame are standing at the ground. For them the train is moving with a
velocity v towards the positive x axis. That means all three of our observers P, A and B has velocities
v with respect to all observers in the reference frame S0 . From the reference frame S0 , the world
lines of P, A and B will be straight lines with slopes tan 1 cv . The world line of the light rays would
be lines with ±45 orientation with the x axis. The events EA and EB can be found by getting the
coordinates of the intersection of the world lines of the light rays with the observers A and B. Clearly,
these two events now would have different time coordinates! These events are no more simultaneous
when recorded the observers in the reference frame S0 .
Our conclusion from this thought experiment is the following. The events happening at different
points in space can be simultaneous from a given reference frame, but they are not necessarily
simultaneous in all other inertial reference frames. Clearly, this demands a change in the Galilean
transformation rules, we can not any more set t = t 0 for any two reference frames.
Problem 1.2 Calculate the slope of the line joining the events EA and EB in the reference frame S0 .
Answer: The slope will be vc .
Problem 1.3 Draw the world line of a cannon ball moving under influence of gravity. Note that for
the cannon ball the motion is in two dimension.
12 Chapter 1. Simultaneity

Fun Facts
We realized that there can be events which may happen at the same time in one reference
frame but does not in other reference frames. For example, in the thought experiment with
the train cart, the event EA happened earlier compared to the event EB in the reference frame
S0 . If the train was moving in the ve x direction then the event EB will appear to happen
earlier. Hence, there are events for which the order of them happening earlier or later depends
on which reference frame you are observing them from. One of these events, then can not
be the cause and the other be the effect! The cause and effect relationship does not hold for
these types of events! We shall discuss this more in the next chapter.
 2. Lorentz Transformations: Space-Time

2.1 Lorentz Transformation Relations

Let us start by writing the Galilean Transformation Rules first:

t 0 = t, (2.1)
0
x = x vt,
0
y = y,
0
z = z.
We have already observed that these are not compatible with the thought experiment discussed in the
last section. Einstein used the postulates of special relativity to modify these transformation rules. It
turned out that the transformation rules he got just by using the two postulates are identical to what
Hendrik Lorentz 1 had already assumed to explain a set of other experiments. In this course we shall
not derive these rules, we shall just write these down here and see it’s consequences.
ct 0 = [ct b x]g, (2.2)
x 0
= [x b ct]g,
0
y = y,
0
z = z,
v 1
where, b= , g=q .
c 1 v2
c2

We shall see that these transformation relations gives rise to many interesting consequences. But
before we take these as granted, we should first check if these relations approximate to the Galilean
Transformation relations when v << c, that is in the condition Galileo derived those doing real
experiments.
1 Hendrik Lorentz received Nobel Prize in Physics in the year 1902 for his explanation of Zeeman effect with Piieter
Zeeman.
 14 Chapter 2. Lorentz Transformations: Space-Time

At the limit v << c, g ! 1 and b ! 0. A little inspection will tell you that indeed in this limit
the Lorentz Transformations approximates to the Galilean Transformation rules.

2.2 Concept of Interval

Let us define a function of the four variables ct, x, y, z that defines an event. We call this function the
interval and write it as s(ct, x, y, z). The interval function is defined such that for any event happening
at the space-time point [ct, x, y, z] (note that we have changed the order of writing the variables),

s2 = c2t 2 x2 y2 z2 . (2.3)

Well, for simplicity again, let us consider only one spacial dimension and not use the y and z
coordinates from now on. All interesting concepts we develop here also holds if you take all three
spacial dimensions. We can also define interval between two events E1 = [ct1 , x1 ] and E2 = [ct2 , x2 ]
as

s212 = c2 (t2 t1 )2 (x2 x1 )2 . (2.4)

Clearly, in view of equation (2.4), equation (2.3) is the case when one of the events is at the origin
of the space-time diagram. We can, in fact, define infinitesimal intervals for two events separated
infinitesimally as

ds2 = c2 dt 2 dx2 . (2.5)

Interval is a measure of distance between two space time points. This concept will be more clear if
we consider the question that how the interval function transforms when we go from one inertial
reference frame to other.
Let us consider an event E that happens at the space-time point E = [ct, x] in the reference frame S,
the corresponding interval is s(ct, x). The same event happens at the space-time point E = [ct 0 , x0 ]
and the corresponding interval is s0 (ct 0 , x0 ). These are given as
s2 = c2t 2 x2 , (2.6)
0 2 0 2 0 2
(s ) = (ct ) (x ) .
We can now use the Lorentz transformation equation to see how the quantities s and s0 are related.
To do this we write ct 0 and x0 in the second equation of 2.6 in terms of ct and x using Lorentz
Transformation equations 2.2. If you do the exercise you will see that s = s0 , i.e, the value of the
interval between two events does not change in Lorentz Transformation.

2.2.1 Light Like Intervals

Let us consider a few example of intervals. Do you remember the thought experiment we did in the
last chapter ? Let us consider the pair of events we called EP = [0, x0 ] and EA = [L, L + x0 ]. Those
were the events of the light rays exiting the torch of our observer P and reaching our observer A. The
world line of the light rays have a slope of 45 or 135 . Let us calculate the interval sPA between
these two events.

s2PA = (c2 ⇥ 0 L)2 (x0 L x0 )2 = 0. (2.7)

 2.2 Concept of Interval 15

Figure 2.1: Examples of Null, Time-like and Space-like Intervals. The dashed lines corresponds
to the world line of light ray or particle along which the intervals are taken. Notice that for the
space-like interval the world line is not drawn since the two events there can not lie in a particle
trajectory.

Calculate the interval between the events EP and EB in our example, you will get zero again. The
reason is that these pair of events are on the world line of light rays. All such events that lie in the
world line of a light ray would be separated by a zero interval. We call these intervals light like
intervals or null intervals.
Can you see why the light like intervals are necessarily null intervals? To see this, let us consider the
definition of the interval as in equation (2.4). We rewrite it as
s212 = c2 (t2 t1 )2 (x2 x1 )2 (2.8)
" ✓ ◆ #
2 2 x2 x1 2
= (t2 t2 ) c
t2 t1
⇥ ⇤
= (t2 t1 )2 c2 v2 .
Here we have used the simple fact that (x2 x1 )/(t2 t1 ) = v, the velocity along the particular world
line in which the two events of the interval lies. But, here the world line is that of light, right? Hence,
for this case, the velocity v = c and hence s12 = 0.
Fun Facts
Here we used the Lorentz Transformation relations to demonstrate that the interval between
two events are invariant or constant under Lorentz Transformation. In fact, the actual story
is other way round. Using the postulates of special relativity you can show directly that
the interval is an invarient. Then this property of the transformation can be used to arrive
at the Lorentz Transformation. Doing this is just beyond the scope of this course. If you
are interested, you can have a look at the book Classical Theory of Fields by Landau and
Lifshitz.

2.2.2 Time Like Intervals

Certainly, it is not necessary that all the intervals between pair of events will be on the world line of
light. In all other cases, the world line will have a slope of tan 1 cv with v 6= c. Hence the intervals
16 Chapter 2. Lorentz Transformations: Space-Time

Figure 2.2: World line of a particle seen from three different reference frames. In the left most panel
the velocity of the particle is zero, where in the middle and right panel panel the velocities are v1
and v2 respectively with v2 > v1 > 0. See the slope of the world line reduces as the velocity of the
particle as seen from a particular reference frame increases.

between a pair of points in such world lines will not be zero.

Let us concentrate our attention to the world lines of particles. If we choose a reference frame say
S0 , where the particle is at rest at x0 , the world line is a vertical line passing through the space-time
point [0, x0 ]. If we observe the same particle from a reference frame S1 where the particle is moving
with a velocity v1 along positive x axis, it has a world line with a slope tan 1 vc1 with the positive x
axis. If we go to a second reference frame s2 , where the particle has a velocity v2 with respect to the
positive x axis, the world line will have a slope tan 1 vc2 . Moreover, if v2 > v1 , the slope of the world
line of the particle as seen from reference frame S2 will be less compared to that in S1 . If we go to a
reference frame where the velocity of the particle is more, the slope of it’s world line will be lesser
(see figure 2.2. In the reference frame where the velocity is zero, the slope is infinite!
So, as we go to reference frames where the velocity of the particle is higher and higher, the slope of
the world line decreases. Let us now ask the following question: A particle is at rest in one reference
frame. Can we go to a reference frame, where the slope of the world line of that particle is one, or
even less than one ? To see this let us take help from the fact that the interval between any pair of
events is independent of the reference frames.
In the reference frame where our particle is at rest, v = 0 and equation (2.8) tells us that the value of
s212 = (t2 t1 )2 c2 > 0. Since the value
⇥ of ⇤s12 can not change by changing the reference frame, from all
reference frames s12 = (t2 t1 )2 c2 v2 > 0, which implies that v < c. In words, for such a particle
we can not find a reference frame where the particle will move at a speed exceeding the speed of light!

Fun Facts
Now, we started this discussion with trajectory of an arbitraty particle. A particle is always at
rest in one reference frame. The discussion leads us to the conclusion that it can not travel
at a speed faster than the speed of light. If it is true for an arbitrary particle, it is then true
for any particle, right? That leads us to the conclusion that: Nothing can travel at a speed
faster than that of light. That is an amazing consequence of the special relativity, is not it?
 2.2 Concept of Interval 17

The intervals which are defined by a pair of events on the trajectory of a particle have a special
property. Let us take an example to understand this. Say Jay throws a ball to Jaya. Assume for the
sake of science that the ball travels at a constant velocity v. The events of throwing the ball and
receiving it are two events in the trajectory of the ball. If you know the velocity of the ball, you will
be able to predict (means tell the position and time coordinate of) all the events in the world line of
the ball. These events actually depends on one another. Every event that has happened earlier is a
cause of all the events happen afterwards. These events are said to be related by cause and effect
relations. The interval is causal. The interval is also such that the time part of the interval c(t2 t1 )
is greater than the space part x2 x1 (see figure 2.1). These intervals with s212 > 0 are called time like
interval. For any such intervals, we can always find a reference frame where the two events happen
at the same point in space. Like for the ball that Jay threw to Jaya, the two events happen exactly at
the same point in space in the frame where the ball is at rest.

2.2.3 Space Like Intervals

Let us now consider two events happening simultaneously at two different positions in a reference
frame S0 . How is the interval for such a pair of events? A little inspection will tell you, for these
events, s212 = (x2 x1 )2 < 0. What happens if we measure the intervals between the same pair of
events from another inertial reference frame? Well, the interval is always constant, in all reference
frames s212 < 0 for such events.
We have seen in the previous chapter that such events which are simultaneous in one reference frame,
would lie in a line with slope of tan 1 vc when observed by observer of another reference frame who
has a velocity v with S0 . Alternatively, for any two events happening in a reference frame when the
interval square for them is negative, we can always find a reference frame where they happen at the
same time. The interval between such events with s212 < 0 are called space like events.
Can we have a particle world line separated by a pair of space like events ? Well, again a little
inspection with equation (2.8) shows that such a particle would need to travel with a velocity v > c,.
This, we have already seen, is not allowed. Since a pair of space like events can not be two events in
the world line of a particle, then one of them can not be a cause for the other. Space-like events do
not have any causal relationship.
 3. Lorentz Transformation: Consequences

To remind you again here are the Lorentz Transformation relations. Suppose that the coordinates
of an event measured from a reference frame S are [ct, x, y, z]. The coordinates of the same event
measured by the observers of the reference frame S0 are [ct 0 , x0 , y0 , z0 ]. All observers in the reference
frame S0 have a relative velocity of v along the x axis with respect to any observer in the reference
frame S.

ct 0 = [ct b x]g, (3.1)

x 0
= [x b ct]g,
0
y = y,
0
z = z,
v 1
where, b= , g=p .
c 1 b2
Can you guess what would be the inverse transformation relations, i.e, the unprimed coordinates
expressed in terms of the primed? Here is the aswer:

ct = [ct 0 + b x0 ]g, (3.2)

x = [x + b ct ]g,
0 0

y = y0 ,
z = z0 .

We have already experienced some non-trivial consequences of the two postulates of special relativity
and Lorentz Transformation relations. We have seen how two events separated by space-like intervals
can never be related by the cause-effect relationship. There can always be a reference frame where
these two events happen simultaneously, but in all other reference frames they are not simultaneous.
We have also seen that the speed of any particle can not exceed the speed of light. We have partitioned
 3.1 Time Dilation 19

Figure 3.1: Two events EA and EB , as seen in the reference frame S happen at the same spacial point
x0 . They happen at different spacial points in the reference frame S0 .

the whole space time in causal and non-causal parts. In this chapter, we shall investigate yet more
bizarre consequences of the Lorentz Transformation.

3.1 Time Dilation

We have already seen in the thought experiment involving the train cart that the difference in time
between two events can change if we measure them from different inertial reference frames. Here
we shall put that into more qualitative way.
Let us consider the world line of an observer who is a part of the reference frame S. Since he is in
the reference frame S, his position x0 , by definition is fixed in S. If we consider any two events in his
world line, they all happen at his position. All this events are measured in his clock, in a single clock.
If two such events are EA = [ctA , x0 ] and EB = [ctB , x0 ] (see figure 3.1), then the interval, denoted by
sAB , between them is given by

s2AB = c2 (tA tB )2 , (3.3)

which is a time like interval and s2AB > 0. The difference in time between these two events is
dt = (tA tB ). Interestingly, this time interval is special as it is measured by the same clock. Such
time intervals between events happening at the same point in space and can be measured by the clock
of the same observer is called the proper time interval. These are the time intervals between events
that happen in the world line of a particle at rest. Let us see how the observers in a reference frame
S0 would measure of these two events. The observers in the reference frame S0 are moving with a
velocity v with respect to the observers in S along the positive x axis. The world line joining these
two events in the reference S0 have a slope of tan 1 cv . Clearly, the events EA and EB do not happen
at the same value of x in the reference frame S0 .
Let us say that the observers in S0 measure the two events as EA = [ctA0 , xA0 ] and EB = [ctB0 , xB0 ]. See
that the spacial coordinates are taken to be different. Now in this reference frame, the interval
20 Chapter 3. Lorentz Transformation: Consequences

between these two events will be defined as

s2AB = c2 (tA0 tB0 )2 (xA0 xB0 )2 . (3.4)

Clearly, the time difference between these two events measured in this reference frame are dt 0 =
(tA0 tB0 ). As we already know that the interval is invariant, we can write

c2 (tA tB )2 = c2 (tA0 t 0 )2 (xA0 xB0 )2 (3.5)

B ⇣ 0 0 ⌘2
x x
) (tA tB )2 = (tA0 tB0 )2 1 c12 tA0 t 0B
A B
h 2
i
) (dt)2 = (dt 0 )2 1 vc2

) dt 0 = p dt = dt g , b = vc , g = p 1 .
1 b2 1 b2

This is an interesting result. We have here b = vc < 1 and hence g > 1. Hence, the time interval
measured between any two events in the world line of a particle in motion is always larger compared
to the proper time interval between the same events. This phenomena is called the time dilation.
Here we have found the relation for the time dilation in equation (3.5) by using the fact that the
interval is constant across all reference frames. We may also use Lorentz Transformation rules to
arrive at the same equations. Let us try that out here.
The events as seen from the reference frame S were EA = [ctA , x0 ] and EB = [ctB , x0 ] and the time
interval dt = (tA tB ). The same events as measured by the observers in reference frame S0 were
EA = [ctA0 , xA0 ] and EB = [ctB0 , xB0 ] and the time interval as measured in the frame S0 is dt 0 = (tA0 tB0 .
Using the Lorentz Transformation relations we get

ctA0 = (ctA b x0 )g, (3.6)

ctB0 = (ctB b x0 )g
) tA0 tB0 = (tA tB )g
) dt 0 = dtg,

the same relation we got before.

 3.2 Length Contraction 21

Figure 3.2: In the reference frame S0 the rod is moving with a velocity v. Hence to measure the
length of the rod location of the two ends of it are marked at the same time t00 .

Fun Facts
• There is a subatomic particle called muon. This particle has very small life time,
about two microseconds. This means, whenever such particles are created, in two
microseconds they break into one electrons and two neutrinos (mostly). These particles
are always created in the upper earth atmospheres from cosmic rays. They travel very
fast, very close to the speed of light. However, even at that speed, at ⇠ 3 ⇥ 108 m sec 1 ,
they can travel only 3 ⇥ 108 ⇥ 2 ⇥ 10 6 = 600 m. Yet these particles can be seen at
lower atmosphere of the earth. By some mysterious means they travel a few kilometers
!
The mystery with muon is no more a mystery once you know about the time dilation.
The life time of the particle as mentioned above is the proper time interval between it’s
creation and disintegration. From the perspective of an observer at earth, the same time
interval is larger by a factor of g as given in equation (3.5). If the muons are travelling
at a speed about 95% that of light, then you can calculate that the time between their
creation and decay would be about ten times the proper time interval and hence the
muons will travel a distance of about 6 km before they disintegrates.

3.2 Length Contraction

Just like the time interval between a pair of events measured in different inertial reference frames are
different, so are the spacial separations. This gives rise to a bizarre phenomena. If you measure and
find the length of a rod in a reference frame, it is not necessary that the length as measured in all
inertial reference frames will be the same. Let us examine this now using Lorentz Transformation
relations.
Let us first ask the following question: How do we measure the length of an object?. It is obvious
22 Chapter 3. Lorentz Transformation: Consequences

that we have to get the spacial coordinates at the ends of the rod and then take their difference. But
is it that we can do these two measurements at any time we wish? Well, the answer is no! Imagine
that the rod is moving towards positive x axis in a reference frame. You do the measurement of the
position of the left end of the rod at a time t1 and then wait for some time and measure the position
of the right end at t2 . Well, the rod has moved towards the right side quite a bit in between and the
difference of readings between the two positions at the left and right end of he rod will certainly be
more than the actual length of the rod. Clearly, to get the correct length, we would have to measure
the two ends of the rods at the same time.

There is, however, a very special reference frame, where the two measurements can be done at
any time. This is the reference frame where the rod is at rest. The rod does not move, so are it’s two
ends. Hence, the measurement of the position of the two ends can be done at any time, we shall get
the same length when we take their difference. This reference frame is special and the measured
length of the rod in this spacial reference frame is called the proper length.

Let us consider the reference frame S0 where the rod is moving with a velocity v along the positive
x direction. The two ends of the rod are measured at the same time and the corresponding events
of measurements are given as EL = [ct00 , xL0 ] and ER = [ct00 , xR0 ] (see figure 3.2) for the measurements
of the left and right ends of the rod. Hence, the length of the rod in S0 is L0 = (xR0 xL0 ). The same
events are given by EL = [ctL , xL ] and ER = [ctR , xR ] in the reference frame S where the rod is at rest.
As discussed earlier, the proper length of the rod will be then L0 = (xR xL ). Using the Lorentz
transformation relations we can write:

xR = (xR0 + b ct00 )g (3.7)

xL = (xL0 + b ct00 )g
xR xL = (xR0 xL0 )g

p
) L0 = L0 [1 b 2 ] = L0 /g

Remember that b = v/c < 1 and hence g > 1. Therefore, the length of a rod as measured in any
reference frame where it is in motion is always shorter than it’s proper length. This phenomena is
called the length contraction.
 3.3 Addition of velocities 23

Fun Facts
• Note that the length contraction only happens along the motion or velocity. Since the
Lorentz transformation for the other two spacial direction is such that they remain
unchanged, the length contraction does not happen along those directions.
• You have studied the rigid body dynamics. If we have a solid object which has very
large number of constituent molecules, it is hard to do classical mechanics with such an
object. The number of coordinates needed to specify the motion of all the molecules in
a solid body is so many that it is impossible to solve so many Euler-Lagrange equations.
We devised a clever technique though. We said, we define these bodies to be rigid, any
two molecules in them have a fixed distance between them. So, we may only use six
coordinates to describe it’s motion, three for the position of the centre of mass and
three for the orientation about the centre of mass. That was a very convenient trick.
Think what happens once you have the length contraction as a consequence of the
Lorentz transformation. You can no longer have the distance between the molecules in
the body fixed once you change the reference frame, there can not be any rigid body in
special relativity. The concept is not useful !

Problem 3.1 Here is a question for you now. Assume that there is a tunnel of (proper) length L with
two doors at the end of the tunnels. There is a train trac through the tunnel. A train, slightly bigger
in (proper) length LT , is passing by with speed near to the speed of light. But for the observers in the
reference frame of the train track, the train is just small enough (because of length contraction) to fit
inside the tunnel momentarily. The gate men at the two ends of the tunnel momentarily closes the
gates just when the train is inside the tunnel and then open it to make it pass. Length contraction
makes it possible, though the tunnel has a proper length smaller than the proper length of the train.
Wait a second ! What happens for the observers who are at the reference frame of the train ? For
them, definitely the length of the train is larger than the tunnel ! Then how come the train fit inside
the tunnel when the two gates are closed ? Use your trick of space time diagram from the two
reference frames to resolve this puzzle.
Problem 3.2 As we did for the time dilation case, the conclusion we will get here as length
contraction can be arrived at just from the notion of intervals. Try it yourself.

3.3 Addition of velocities

When we discussed the experiment by Michelson and Morley to find the speed of the earth v with
respect to the fixed ether frame, we used Galilean transformation relations to find the speed of light
from the earth as compared to it’s value of c from the ether. When the velocity of light is along the
motion of the earth through the ether, the speed was taken to be c v, for example. Now we know
that the speed of light is a constant irrespective of the observer’s and source’s motion. Hence, let us
re examine the laws of addition of velocities.
Here we are asking the following question. In the reference frame S a particle is moving with
a velocity V along the positive x axis. The observers in the reference frame S0 are moving with a
velocity v with respect to the observers in the reference frame S. What is the velocity of the particle
as measured by the observers in the reference frame S0 ? Let us denote this velocity by V 0 . Our goal
here is to find a relation between V and V 0 .
24 Chapter 3. Lorentz Transformation: Consequences

Let us first ask the question, what is the velocity of the particle in terms of the coordinates ?
Velocity of the particle from any frame is the rate of change of its position in unit change in time.
dx0
That is, the velocity can be defined as V = dx 0
dt and V = dt 0 .
Using the Lorentz transformation relations we may write:
cdt 0 = [cdt b dx]g, (3.8)
dx 0
= [dx b cdt]g,
0
dy = dy,
0
dz = dz.
Hence, the x component of the velocity in the reference frame S0 can be written as:
dx0
Vx0 = (3.9)
dt 0
(dx b cdt)g
=
(cdt b dx)g
dx
cdt b
= b dx
1 c dt
Vx v
) Vx0 = vVx .
1 c2

In general, a particle can move along arbitrary direction in a reference frame. That is, it can have all
three components of velocity in the reference frame S. Using the same calculations for the y and z
component of velocities we get the velocity transformation rules:
Vx v
Vx0 = vVx (3.10)
1 c2
Vy
Vy0 = vVx
1 c2
g
Vz
Vz0 = vVx .
1 c2
g

Problem 3.3 Can we use this for the case of light now ? There is a reference frame S where we
send a light ray along the positive x direction with the speed c. What is the velocity of the light ray
from other reference frames ?
Solution: It is clear that in other reference frames the speed along y and z directions be zero (see
equation 3.10). Along the x direction the velocity will be
c v c v
c0 = cv =c = c. (3.11)
1 c2
c v

See, Lorentz transformation preserves the speed of the light !

Problem 3.4 Consider in a reference frame S a particle is moving along negative x direction with a
velocity 3c/4. The observers in another reference frame S0 are moving with a velocity 3c/4 along
the positive x direction with respect to the observers in S. Use the velocity addition formula to find
the velocity of the particle as seen from the reference frame S0 .
3.3 Addition of velocities 25

Problem 3.5 In a reference frame S a particle is observed to have a velocity 2c x̂ + 3c ŷ. Use the
velocity addition formula the velocity of the velocity of the particle as seen from a reference frame
S0 . The observers in S0 are moving with a velocity 3c/4 along the positivex direction with respect to
the observers in S.
 4. Dynamics of a free particle

We have seen that Lorentz transformation relations that follow from the postulates of special relativity
require modification of the Galilean rules of kinematics, the addition of velocities. In this chapter we
shall investigate the dynamics of a free particle and eventually see the most famous equation of all
time.
Let me first keep it clear that there can be many approaches taken to do what we shall be going
through in this chapter. We shall take one of the approaches. Between these different approaches, the
definition of different dynamical entities like mass, momentum and energy changes. Though these
different approaches to relativistic dynamics are identical, all of them can not be used for all the
systems. The approach we shall take here, however, can be used to solve a large class of problems.
In this approach, we use the mass of a particle to define a neutral particle. If the particle you are
interested in is a charged particle, you also have to specify it’s charge. These characteristics can not
change even if we measure the dynamics of the particle from different inertial (or any) reference
frame. 1 We shall define the momentum of the particle as the canonical momentum as we have
learned in the Lagrangian Mechanics. Similarly, we will use the Hamiltonian for a free particle to
calculate the energy of the particle.

4.1 Lagrangian of a free particle

We would first write the Lagrangian of a free particle and use the Lagrangian mechanics we have
learnt earlier to proceed to find the dynamics of a free particle. How do we write the Lagrangian of a
system? Well in Classical Mechanics we have used the approach where the Lagrangian is considered
as the difference of the total kinetic energy and total potential energy. But have you ever thought why
1 You may be vaguely familiar with another approach to relativistic dynamics and have heard the term ‘rest mass’. In
that approach to the relativistic dynamics the mass of a particle changes in different reference frame and the mass defined
in the reference frame where it is at rest is called the ‘rest mass’. We are not using this approach. What we define as mass
of the particle is essentially the ‘rest mass’ as defined in the other approach.
 4.2 Relativistic Momentum 27

is the Lagrangian of a system have this form? This question has not an easy answer. In a nutshell, if
we write the Lagrangian in this way, we get the right equation of motion. Remember the Lagrangian
L is a function of the coordinate x, the time derivative of it ẋ and time t. The functional form of the
Lagrangian for a given system is always to be written down only using the first principals for the
problem.
To analyze the dynamics of a particle moving with relativistic speed we can use the postulates of
special relativity and Newton’s first law of mechanics to get the functional form of the Lagrangian.
Here, we do not trace the arguments to write this Lagrangian, but directly write it down. Lagrangian
for a free particle moving in relativistic speed is
r
2 v2
L(x, y, z, ẋ, ẏ, ż) = mc 1 , (4.1)
c2

where the velocity v is the magnitude of the velocity of the particle given as ~v = ẋx̂ + ẏŷ + żẑ. Since
it is a free particle, certainly it will move at a constant speed (Newton’s first law of mechanics). This
Lagrangian has the dimension of energy as expected. If we consider the non-relativistic limit, when
the velocity of the particle v << c, we get

2 v2 1
LNR = mc 1 2
= mc2 + mv2 . (4.2)
2c 2

This has the exact form of a free particle Lagrangian (we knew from the nonrelativistic physics)
with an extra constant term of mc2 . Well, we have already learnt that adding a constant term in the
Lagrangian does not do anything to the dynamics of the particle. Hence, in the non-relativistic limit
this Lagrangian gives rise to the exact same dynamics we are familiar with.

4.2 Relativistic Momentum

Let us calculate the canonical momentum using the above Lagrangian. The x component of the
canonical momentum will be

∂L
px = (4.3)
∂ ẋ
1 ẋ 1
= mc2 ( 2 2 ) q
2 c 1 v2
c2
= mẋg

Clearly, if we consider all the three components of the momentum it can be defined as

~p = m~vg. (4.4)

Do you remember that g is always greater than or equal to one ? That means the momentum for
a particle does not linearly increase with it’s velocity. As the velocity increases the momentum
increases faster.
 28 Chapter 4. Dynamics of a free particle

4.3 Relativistic Energy

Once we know of the Lagrangian and the canonical momentum, it is straight forward to calculate the
energy for the particle. The Hamiltonian is given as
H = ~p.~v L
= m~vg.~v ( mc2 /g)
 ✓ ◆
2 2 v2
= mg v + c 1
c2
 2
v v2
= mc2 g 2 + 1
c c2
= mc2 g. (4.5)
As we already know that the energy of a free particle is equal to it’s Hamiltonian, we get

E = mc2 g. (4.6)

Well, what happens for the particle at rest ? This particle has a velocity v = 0 and hence g = 1 and
you see

E0 = mc2 , (4.7)

where we see that a free particle at rest has some minimum energy E0 , same as its mass times the
square of the speed of light. This is the most famous equation of all time.

4.4 Relativistic Momentum-energy relation

In the usual classical mechanics the energy of a free particle is given as

p2
E= . (4.8)
2m
Also in classical mechanics a free particle at rest (i.e p = 0) has zero energy. We have seen already
that a particle at rest has a non-zero energy of mc2 . Since the definition of the energy and momentum
is different for the particles moving in relativistic speed, we need to investigate the energy-momentum
relation for them. Let us calculate the following:
E2 p2 c2 = m2 c4 g 2 m2 v2 c2 g 2 (4.9)

v2
= m2 c4 g 2 1
c2
= m2 c4 g 2 /g 2
Hence, we get the relation between the relativistic energy and momentum as

E2 p2 c2 = m2 c4 . (4.10)

Let us try to examine this relation in a little more detail. In the right hand side, we have the mass
of the particle m (or the rest mass if you are following the other approach) and the speed of light c.
 4.5 Final Notes 29

Both of these are constants, hence the left hand side is a constant. Earlier we saw that the energy and
the momentum of the particle changes as you go from one reference frame to other, however, it is
clear that a combination of the E and p remain unchanged.
An interesting consequence of the above relation is that we now can have particles which has no
mass but still have momentum ! If the particle has an energy E, then the momentum will be simply
given as p = E/c. The light quantum of Einstein, which we now call photon is an example of such a
mass-less particle.
Fun Fact
You may be thinking that having a momentum for the light is just a theoretical concept. It is
as real as it can be. A consequence of light having momentum is that if you stop the light
or it bounces back from a surface the momentum changes. Change in momentum is force.
This is an well known measurable effect for the artificial satellites. Radiation from the sun
hits the solar panels of the artificial satellites. Every now and then the rate of change in the
momentum is asymmetric in the two halves of the panels and that gives rise to a torque. The
satellites start to rotate. The satellites, hence, requires to have fuel to counteract this rotation.

4.5 Final Notes

Special relativity has been a revolution in physics. It not only fixed the quarrel between Maxwellian
electrodynamics and Newtonian Mechanics, it also gave rise to concepts of spin of the electron,
negative energy states of system etc. The de Broglie hypothesis we have learned earlier is a result of
a relativistic viewpoint. Over the last hundred years, many experiments are performed to verify the
effect of the special relativity and for all of them, the theory emerged victorious.