Circles, Ellipses, and Elliptic Integrals
Archimedes
In his work Κύκλου Μέτρησις (Measurement of a Circle), Archimedes
calculates successive approximations to
which is equal to , by using reduction formulas relating the perimeters of regular
inscribed and circumscribed
-gons to those of the corresponding regular
-gons of the circle of radius one.
Why does the integral appear?
Consider the upper semicircle
To compute its arc length, partition the interval into subintervals
and approximate the semicircle by the polygonal path joining the
points
for .
By the Pythagorean Theorem, the length of the -th segment is
where and
. Assuming a uniform partition,
is constant, and we may
write
By the Mean Value Theorem, there exists a point
such that
Hence the length of the segment
becomes
Summing the lengths of all segments gives the approximation
which is the Riemann sum for
Since
the arc length of the upper semicircle is
The integrand
is continuous on but it blows up at
. So we interpret the integral as an
improper integral:
For , the integrand is continuous and
Therefore,
Since is continuous at
Hence
Greek mathematicians carried out such computations using only a
ruler and a compass.
What follows is not the method Archimedes himself used to
approximate the circumference of the unit circle. Rather, it is an analytical
reconstruction of the ideas underlying his procedure.
The reduction formula that relates
the circumscribed -gons to the corresponding
-gons follows directly from the figure above. Indeed,
and
Since, in a triangle, the bisector of an interior angle divides the opposite side into segments whose lengths are proportional to the lengths of the adjacent sides (see Basic Geometry in The Riemann Hypothesis Revealed), we have
From this we get:
or
For the reduction formula that
relates the inscribed -gons to the corresponding
-gons (see the following figure) we have
and
Triangles ,
and
are obviously similar.
Therefore
Adding yields
Using Pythagoras’ Theorem
Since ,
and
, we obtain
For convenience set and
. So
or
Thus,
the side lengths of the circumscribed and inscribed polygons satisfy the
reduction formulas
and
for
.
Now let
be the perimeters of the inscribed
and circumscribed polygons.
Obviously
If is the apothem corresponding
to the side of the regular
-gon then from Pythagoras’ Theorem
From
we get
hence
Thus,
Additionally,
and for any there exists an
-gon such that
The side length of the inscribed -gon is
and
. Also, the side length of the circumscribed
-gon is
. Therefore (this can be proved purely from the geometry of the
inscribed and circumscribed polygons, with no trigonometry at all)
and
So
Since
the squeeze theorem implies
This is essentially Archimedes’ use
of the Squeeze Theorem. Consequently,
The Man Who Knew Infinity
In the movie The Man Who Knew Infinity (2015), there is a memorable
classroom scene at Cambridge in which Ramanujan, having neglected to take notes
during a lecture, is challenged by Professor Howard to contribute to the
evaluation the integral:
This integral is known as the complete
elliptic integral of the first kind. Elliptic integrals owe their name to
their historical appearance in the problem of determining the arc length of an
ellipse, although they arise naturally in many other areas of mathematics and
physics.
An ellipse is defined by the major
axis of length and the minor axis of length
. The exact perimeter of an ellipse
is given by the integral
where is the eccentricity of the ellipse.
The integral
is known as the complete elliptic integral of the second
kind.
To see why
we start with a natural
parametrization
because
Use the arc-length formula for a
parametrized curve over the first quadrant
Here,
Therefore
Factoring out :
But
so
Thus,
Infinite Sums
In the movie Ramanujan immediately
writes on the blackboard:
astonishing both the professor and
the class. The scene is intended to showcase the exceptional mathematical
insight for which Ramanujan became legendary.
To prove this result, we follow
the approach presented in (Almost) Impossible Integrals, Sums, and Series
by Cornel Ioan Vălean. We begin with the following auxiliary integral:
We have:
Now set
Set
Now take
Set in
. Then
Thus,
Expand
so
Now use the well-known Wallis’ integral
It follows that
To find a similar formula for we write the integrand as
and use the binomial series expansion for with
,
:
where .
Set . Then
Assuming term wise integration is justified (it is ), we get
We have
Multiplying by
so
because
Thus,
Apply this formula to find the perimeter of the ellipsis:
Epilogue
When the subject is rich in geometry, I find myself
especially excited. Even more so when I can build my ideas around the geometric
intuition of Archimedes. There is something deeply satisfying about seeing an
abstract mathematical argument come alive through geometry.
But there is another important element: the reader and I need
to have some common ground. For my essays, that common ground is provided by my
book, The Riemann Hypothesis Revealed. Everything I use in these essays can, of
course, be found elsewhere, and I make no claim otherwise. But if you are
looking for a single book that brings together all the mathematical tools and
background needed for these ideas, this is the book for you.
The Riemann Hypothesis Revealed is available through Amazon
marketplaces worldwide.
Comments
Post a Comment