The Gamma Function, Factorials and

the Volumes of n-Balls

Wrapping up First Year Calculus in an n-Ball

Bruce Cohen and David Sklar

California Math Council
Asilomar Conference
December 1, 2001

How to Reach us

Bruce Cohen
Lowell High School
San Francisco, CA 94132
bic@cgl.ucsf.edu
http://www.cgl.ucsf.edu/home/bic

David Sklar
Sola Optical USA
Petaluma, CA 94955
dsklar46@yahoo.com

Our goal is to demonstrate to the students that the

mathematics they learned this year has provided them
with a very powerful problem solving tool.
We use the tools of first year calculus and some
geometric intuition to derive a surprisingly simple
formula for the volume of an n-dimensional ball.
Along the way we learn about the gamma function,
the factorial function for non-integer values and some
interesting trigonometric integrals.

What is Calculus?
I begin each school year with the idea that Calculus can be
viewed from the perspective of three problems.
1. We know the slope of a line.
What is the slope of a curve at a point?
2. We know the area under a line segment.
What is the area under a curve?
3. We can compute values for functions that contain only
addition, subtraction, multiplication, and division. How can
we compute values for functions that use other operations?

Side Issues in Calculus

Interpolating functions to increase the domain.
e.g.

f x

integers

2
rationals

reals

New operations for function machines

Arithmetic: , ,g,
Powers and roots
Exponential and log
Trig

Using Integrals:
Accumulation Functions
x

A( x ) f (t )dt
a

Other parameters

P ( x ) f ( x, t )dt
a

Interpolating functions to increase the domain

Consider n ! n n 1 K 2 1

n 1 ! n 1 K 2 1
we have n ! n n 1 !

Since

Note that nothing in this recursive definition demands

n to be an integer. (except a starting value)
Perhaps we can extend the domain of factorials.

New operations for function machines:

The Gamma Function
We begin with a definition of the gamma function using
one of the powerful new methods calculus provides for
defining functions the improper integral.

x 1 t

t e dt

for x 0

Then some easy exercises to gain familiarity

evaluate

graph f x, t = t e

x 1 t

for x 0, , 1, 2, 3, 4,5
2

The Gamma Function

Evaluate 1

x
1

t x 1e t dt
11 t

t e dt

lim
n 0

lim
n

lim
n

e t dt

1 1
1
0 lim 1 n 1
n
n
e
e
e

Gamma Integrand
x=0

f x, t = t x 1e t

Gamma Integrand
x=1/2

f x, t = t x 1e t

Gamma Integrand
x=1

f x, t = t x 1e t

Gamma Integrand
x=2

f x, t = t x 1e t

Gamma Integrand
x=3

f x, t = t x 1e t

Gamma Integrand
x=4

f x, t = t x 1e t

Gamma Integrand
x=5

f x, t = t x 1e t

Gamma Integrand
x is 0, 1/2, 1, 2, 3, 4, 5

f x, t = t x 1e t

The Gamma Function

Show x 1 x x
x 1

ut

t x e t dt

Use integration by parts.

t

dv e dt

du xt x 1dt
x t

( x 1) t e

v e t

| e xt

x 1

dt

( x 1) 0 xt e dt = x t x 1e t dt
x 1 t

x 1 x x

The Gamma Function

How about n for integer values of n?
x 1

x( x)

(1 1) 1(1) 11 1!

(2 1) 2(2) 2 1 2!

(3 1) 3(3) 3 2! 3!

(4 1) 4 (4) 4 3! 4!

g
g
g

n 1 (n 1) n(n) n (n 1)! n !

The Factorial Function

The result n +1) = n! suggests a way to extend the domain of the
factorial function beyond the non-negative integers.
We define the factorial function for real x > -1 by

x ! x 1
Before finding some new values of the factorial function we
need one special value of the gamma function. We need
to evaluate 1 2 .

Gamma Function

Factorial Function

x ! x 1

x t x 1e t dt
0

3.5

3.5

2.5

2.5

1.5

1.5

0.5

0.5

0.5

1.5

2.5

3.5

-1

-0.5

0.5

1.5

2.5

1
x

t x 1e t dt
1
1
2

Letting u t

e dt

e t dt
du 1 2 t

tu ,
2

dt

u dan t dimulai dari 0 sampai

e dt e
t

u2

2du 2 e
0

u2

du e

u2

du

f x 3 e

x2

2.5

2
1.5
1
0.5
-3

-2

-1

Two Views
A mathematician is one to whom

x2

e dx

is as

obvious as that twice two makes four is to you.

Lord Kelvin
Many things are not accessible to intuition at all, the value
of

x2

dx for instance.
J. E. Littlewood

1 2

u2

du

We have considered two options for trying to get the

students to accept this:
1.

a numerical exploration

2. a proof that follows from the properties of a

sequence of integrals that we investigate later.

Factorials
We now go back to the definition

x ! x 1

Using the recursion x 1 x x we can show that

x ! x 1 x x x x 1 1 x x 1 !
now with x ! x x 1 ! and 1 2 we can evaluate

1 ! , 2 ! ,

2 2

3 ! ,

2

! , 5 !,

2
2

6 ! , ... ,

2

n
2

Factorials
1 ! 1 1 1

2
2
2

2 ! 1 ! 1


2
3 ! 3 3 1 !


2 2 2
4 !

2
5 !

2
6 !

2
7 !

2

2
6

2
7

2 !

2
3 !

2
4 !

2
5 !

2

4
2

2
6
2

7

2

1
2

1 ! 3

2 2
2

1
2

1
3

2 1
5

 n !

2
n

for n even !
2

n 1 n 2 3 2 1

2
2
an ordinary integer factorial

n
for n odd !
2

n 1 n 2

2
2

3
2

1
2

we march nicely down to 1 2 and end with a

n-Space, n-Balls and n-Spheres

Euclidean n-dimensional space is the set of ordered Real n-tuples
together with a Euclidean distance measure.
R n x1 , x2 ,..., xn | x1 , x2 ,..., xn R

with the distance from the origin to x1,..., xn x1 x2 xn

2

We define the n-ball, Bn R , of radius R and its boundary, Sn 1 R by

R x , x ,..., x R

Bn R
Sn 1

x1 , x2 ,..., xn R n | x1 x2 xn R 2
1

| x1 x2 xn

1-ball, 2-ball, 3-ball,

R

x1

B1 R
x2

x3

S1 R

x2

x1

B2 R

x1

B3 R

1-ball, 2-ball, 3-ball,

Some notation then some examples

let: Vn R The "n -volume" of Bn R

An 1 R The "n 1 - volume" of Sn 1 R

Examples
n 1

B1 R x R | x 2 R 2 x R | R x R

R
V1 R 2 R

R
A0 R ?

1-ball, 2-ball, 3-ball,

Examples
n2

B2 R x, y R 2 | x 2 y 2 R 2

S1 R
V2 R R 2
x

B2 R

A1 R 2 R

1-ball, 2-ball, 3-ball,

Computing V2 R using the "disk" method

x2 y2 R2
x, y
r

V 2 r x
r y R2 x2
dV 2 R 2 x 2 dx

V2 R dV 2 R 2 x 2 dx
2

R x dx 4
2

R 2 x 2 dx

using the trigonometric substitution

x R sin , dx R cos d
R 2 x 2 R 2 1 sin 2 R 2 cos2

x goes from 0 to R as goes from 0 to

2

2
0

V2 R 4 R 2 cos2 d 4 R

2 2
0

cos2 d

1-ball, 2-ball, 3-ball,

Examples
n3

B3 R x, y , z R 3 | x 2 y 2 z 2 R 2

4
V3 R R 3
3
A2 R 4 R 2

x
B3 R

1-ball, 2-ball, 3-ball,

Computing V3 R using the disk method

y 2 z2 R2
0, y, z
r

V r 2 y
r z R2 y 2
dV R y dy
2

V3 R dV R 2 y 2 dy
R

y 2 dy

using the trigonometric substitution

y R sin , dy R cos d
R 2 y 2 R 2 1 sin 2 R 2 cos2

y goes from 0 to R as goes from 0 to

2

V3 R 2 R 3 2 cos3 d
0

recall V2 R 4 R

2 2
0

cos2 d

1-ball, 2-ball, 3-ball,

n
0
1
2
3
4
g
g
g

Vn R
?
2R

An 1 R
?
?

R2
4 3
R
3
?

2 R
4 R 2

g
g
g

g
g
g

1-ball, 2-ball, 3-ball, 4-ball,

Examples
n4

B4 R x, y , z, w R 4 | x 2 y 2 z 2 w 2 R 2

V4 R ? R 4
A3 R ?? R 3

1-ball, 2-ball, 3-ball,4-ball,

Computing V4 R using the disk method

When we slice the 4-ball at a fixed value of z

z 2 w2 R 2
0,0, z, w
r
z

x, y , z, w R

| x 2 y 2 z 2 w2 R 2

reduces to

x, y , z, w | x

y 2 w2 R 2 z 2 r 2 .

4 3
** V r z
3
r w R2 z2
4
dV
3

R z
2

dz

1-ball, 2-ball, 3-ball,4-ball,

Computing V4 R using the disk method

z 2 w2 R 2
0,0, z, w
r

4 3
** V r z
3
r w R2 z2
4
dV
3

R z
2

dz

3
4
2
2
V4 R dV
R z dz
R
R 3
3
8 R
2
2

R z dz
0
3
using the trigonometric substitution
z R sin , dz R cos d
R 2 z 2 R 2 1 sin 2 R 2 cos2

z goes from 0 to R as goes from 0 to

2
8 4 2
V4 R R cos4 d
0
3
R

1-ball, 2-ball, 3-ball,4-ball,

Computing V4 R using the shell method

z 2 w2 R 2
R

V4 R dV A3 r dr
0

r
z

V4 R A3 r dr is in the form
0

F x f t dt hence by the
0

** V A3 r r
dV A3 r dr

Fundamental Theorem of Calculus

d
V4 V4 R A3 R
dR

 , n-ball,
Bn R

x1,..., xn 1, xn R n | x12 xn 12 xn 2 R 2

Vn R ? R n
An 1 R ?? R n 1
We can be more explicit
Vn R cn R n
An 1 R Vn R ncn R n 1

1-ball, 2-ball, 3-ball, 4-ball, , n-ball,

n
0
1
2
3
4
g
g
g

Vn R
?
2R

An 1 R
?
2

R2
4 3
R
3
c4 R 4

2 R

4c4 R 3

g
g
g

g
g
g

cn R n

ncn R n 1

4 R 2

 , n-ball,
again using the disk method

xn

xn xn 1 R
0,...0, xn 1 , xn
r

xn 1

V Vn 1 r xn 1 Vn 1 xn xn 1
r xn R 2 xn 1
dV Vn 1

R x

2
n 1

Vn R dV Vn 1
R

2 Vn 1
0

dx

n 1

R xn 1 dxn 1
2

R xn 1 dxn 1
2

 , n-ball,
R

Vn R dV 2 Vn 1 r dxn 1 2 Vn 1
R

R 2 xn 1 dxn 1

** letting Vn r cn r n , we have Vn 1 r cn 1r n 1 and

R

Vn R 2 Vn 1
0

R x
2

2
n 1

dx

n 1

2cn 1

R xn 1
2

n 1

dxn 1

now using the trigonometric substitution xn 1 R sin implies

dxn 1 R cos d and R xn 1 R 2 1 sin 2 R 2 cos 2 .

2

So Vn R 2cn 1R

n 2
0

cos n d 2cn 1I n R n ,

Hence Vn R cn R n 2cn 1 I n R n ,

cn 2cn 1 I n

2
0

where I n cos n d
which implies

Some Interesting and useful Integrals

Let

2
0

I n cos n d

Show

I 0 , I1 1
2

Show

1
n 1
n 1
n 2
cos

sin

cos

cos
d

n
n

Show

2
0

1
n 2
2
cos n d
cos
d

n 0

Hence

n 1
In
I n 2
n

I n for even and odd n

I0

, I1 1 and

I0
2
1
1
I2 I0
2
2 2
3
3 1
I4 I2
4
4 2 2
I6

I 2k

5
5 3 1
I4
6
6 4 2 2

In

n 1
I n 2
n

I1 1
2
2
I1
3
3
4
4 2
I5 I3
5
5 3
I3

I7

6
6 4 2
I5
7
7 5 3

g
g

g
g

2k 1 2k 3
2k 5
3 1

2k 2 k 1 2 k 2
4 2 2

I 2 k 1

2k 2 k 1 2 k 2 6 4 2

2k 1 2k 1 2 k 3
7 5 3

Vn R cn R n ,

Back to the n-ball

cn 2cn 1 I n also V1 R 2 R

c1 2

c2 2c1 I 2 2 2
4

1
c8
4
4 3 2

4
2
c3 2c2 I 3 2
3
3
1 2
4 3 1
c4 2c3 I 4 2

2
3 4 2 2
8 2
1 2 4 2
c5 2c4 I 5 2

53
2 5 3

25
c9
4
9 7 5 3

4 2 2
c6 2c5 I 6 2

5 3
1 3
c7 2c6 I 7 2
6

5 3 1
1 3

6 4 2 2
3 2
6 4 2
16
3

7 5 3
7 53

1
c10
5
5 4 3 2
g
g
g

n2

if n is even
n !

cn
2

??

if n is odd

???

Back to the n-ball

c1 2
4
c3
3
8 2
c5

53
24
c7
3
7 53
25
c9
4
9 7 5 3
g

cn
n

2 !

if n is even

What about n odd ?

9 7 5 3 1

2 2 2 2 2
4

2 !

g
g

cn
n

2
if n is even or odd
if n is odd so cn
n !
2 !
2
n

n
0
1

Finally

Vn R cn R
n
n

An 1 R n
n

2 !

2 !

n 1

3
4
5
6
g
g
g

Vn R
1
2R

R2
4 3
R
3
2 4
R
2
8 2 5
R
5
3 6
R
6
g
g
g

An 1 R
2
2 R
4 R 2
2 2 R 3
8 2 R 4

3R5
g
g
g

Bibliography
E. Artin, The Gamma Function, Holt, Rinehart and Winston, New York, 1964
( a translation of : E. Artin, Einfuhrung in die Theorie der Gammafunktion,
Leipzig, 1931).
P. J. Davis, Leonhard Eulers Integral: A Historical Profile of the Gamma
Function, Amer. Math. Monthly vol. 70 (1963)
( also in: The Chauvenet Papers, Volume II, M.A.A. 1978 ).
Littlewood, J.E. Newton and the Attraction of the Sphere, Mathematical
Gazette, vol. 63, 1948.
Spivak, M Calculus, Publish or Perish, Inc. Houston, 1994.
Thompson, S.P. Life of Lord Kelvin, Macmillan, London, 1910.