for any even n the maximal length of a centrally symmetric inscribed n-gon in E

having a given point P of E as a vertex is independent of P. Theorem 2 can be

proved and generalized along the same lines. The generalization reads as follows: if
E1 , . . . , Ek are confocal ellipses, then the maximal length of a centrally symmetric
2k-gon (P1 , . . . , Pk , −P1 , . . . , −Pk ) with Pi in Ei , as P2 , . . . , Pk vary with P1 fixed,
is independent of P1 .

REFERENCES

1. M. Berger, Géométrie, vol. 2, Nathan, Paris 1990.

2. J.-M. Richard, Safe domain and elementary geometry, Eur. J. Phys. 25 (2004) 835–844.
3. S. Tabachnikov, Billiards, Panorames et Synthèses 1, Société Mathématique de France, Paris, 1995.
4. , Geometry and Billiards, American Mathematical Society, Providence, RI, 2005.

Collège de France, 3, rue d’Ulm, 75005 Paris, France

[Link]@[Link]
Max-Planck-Institut für Mathematik, Vivatsgasse 7, 53111 Bonn, Germany
zagier@[Link]

An Elementary Proof of the

Wallis Product Formula for pi
Johan Wästlund

1. THE WALLIS PRODUCT FORMULA. In 1655, John Wallis wrote down the
celebrated formula
2 2 4 4 π
· · · ··· = . (1)
1 3 3 5 2
Most textbook proofs of (1) rely on evaluation of some definite integral like
 π/2
(sin x)n d x
0

by repeated partial integration. The topic is usually reserved for more advanced cal-
culus courses. The purpose of this note is to show that (1) can be derived using only
the mathematics taught in elementary school, that is, basic algebra, the Pythagorean
theorem, and the formula π · r 2 for the area of a circle of radius r .
Viggo Brun gives an account of Wallis’s method in [1] (in Norwegian). Yaglom and
Yaglom [2] give a beautiful proof of (1) which avoids integration but uses some quite
sophisticated trigonometric identities.

2. A NUMBER SEQUENCE. We define a sequence of numbers by s1 = 1, and for

n ≥ 2,
3 5 2n − 1
sn = · ··· .
2 4 2n − 2

914 
c THE MATHEMATICAL ASSOCIATION OF AMERICA [Monthly 114
The partial products of (1) with an odd number of factors can be written as

22 · 42 · · · (2n − 2)2 · (2n) 2n

on = = 2, (2)
1 · 3 · · · (2n − 1)
2 2 sn
while those with an even number of factors are of the form
22 · 42 · · · (2n − 2)2 2n − 1
en = = . (3)
1 · 32 · · · (2n − 3)2 · (2n − 1) sn2
Here e1 = 1 should be interpreted as an empty product. Clearly en < en+1 and on >
on+1 , and by comparing (2) and (3) we see that en < on . Therefore we must have

e1 < e2 < e3 < · · · < o3 < o2 < o1 .

Thus if 1 ≤ i ≤ n,
2i
= oi ≥ on
si2
and
2i − 1
= ei ≤ en ,
si2
from which it follows that
2i − 1 2i
≤ si2 ≤ . (4)
en on
It will be convenient to define s0 = 0. Notice that with this definition, (4) holds
also for i = 0. We denote the difference sn+1 − sn by an . Observe that a0 = 1, and for
n ≥ 1,
 
2n + 1 sn 1 3 2n − 1
an = sn+1 − sn = sn −1 = = · ··· .
2n 2n 2 4 2n
We first derive the identity
j +1 i +1
ai a j = ai a j +1 + ai+1 a j . (5)
i + j +1 i + j +1

Proof. After the substitutions

2i + 1
ai+1 = ai
2(i + 1)
and
2j + 1
a j +1 = aj,
2( j + 1)
the right hand side of (5) becomes
 
2j + 1 j +1 2i + 1 i +1
ai a j · + · = ai a j .
2( j + 1) i + j + 1 2(i + 1) i + j + 1

December 2007] NOTES 915

 If we start from a02 and repeatedly apply (5), we obtain the identities

1 = a02 = a0 a1 + a1 a0 = a0 a2 + a12 + a2 a0 = · · ·
= a0 an + a1 an−1 + · · · + an a0 . (6)

Proof. By applying (5) to every term, the sum a0 an−1 + · · · + an−1 a0 becomes
     
1 n−1 2 1
a0 an + a1 an−1 + a1 an−1 + a2 an−2 + · · · + an−1 a1 + an a0 . (7)
n n n n

After collecting terms, this simplifies to a0 an + · · · + an a0 .

3. A GEOMETRIC CONSTRUCTION. We divide the positive quadrant of the x y-

plane into rectangles by drawing the straight lines x = sn and y = sn for all n. Let Ri, j
be the rectangle with lower left corner (si , s j ) and upper right corner (si+1 , s j +1 ). The
area of Ri, j is ai a j . Therefore the identity (6) states that the total area of the rectangles
Ri, j for which i + j = n is 1. We let Pn be the polygonal region consisting of all
rectangles Ri, j for which i + j < n. Hence the area of Pn is n (see Figure 1).

..
.

35/16
R0,3
15/8
R0,2 R1,2
3/2

R0,1 R1,1 R2,1

1

R0,0 R1,0 R2,0 R3,0

0
0 1 3
2
15
8
35
16
···

Figure 1. The region P4 of area 4.

The outer corners of Pn are the points (si , s j ) for which i + j = n + 1 and 1 ≤
i, j ≤ n. By the Pythagorean theorem, the distance of such a point to the origin is

si2 + s 2j .

916 
c THE MATHEMATICAL ASSOCIATION OF AMERICA [Monthly 114
By (4), this is bounded from above by
 
2(i + j) 2(n + 1)
= .
on on

Similarly, the inner corners of Pn are the points (si , s j ) for which i + j = n and 0 ≤
i, j ≤ n. The distance of such a point to the origin is bounded from below by
 
2(i + j − 1) 2(n − 1)
= .
en en
√
Therefore Pn contains √ a quarter circle of radius 2(n − 1)/en , and is contained in
a quarter circle of radius 2(n + 1)/on . Since the area of a quarter circle of radius r
is equal to πr 2 /4 while the area of Pn is n, this leads to the bounds

(n − 1)π (n + 1)π
<n< ,
2en 2on
from which it follows that
(n − 1)π (n + 1)π
< en < on < .
2n 2n
It is now clear that as n → ∞, en and on both approach π/2.

REFERENCES

1. V. Brun, Wallis’s og Brounckers formler for π (in Norwegian), Norsk matematisk tidskrift 33 (1951) 73–
81.
2. A. M. Yaglom and I. M. Yaglom, An elementary derivation of the formulas of Wallis, Leibnitz and Euler
for the number π (in Russian), Uspechi matematiceskich nauk. (N. S.) 57 (1953) 181–187.

Department of Mathematics, Chalmers University of Technology, S-412 96 Göteborg, Sweden

wastlund@[Link]

Automorphisms of Finite Abelian Groups

Christopher J. Hillar and Darren L. Rhea

1. INTRODUCTION. In introductory abstract algebra classes, one typically en-

counters the classification of finite Abelian groups [1]:

Theorem 1.1. Let G be a finite Abelian group. Then G is isomorphic to a product of

groups of the form

H p = Z/ p e1 Z × · · · × Z/ p en Z,

in which p is a prime number and 1 ≤ e1 ≤ · · · ≤ en are positive integers.

December 2007] NOTES 917