*The author uses Newton`s impact method to compute the...*

**The Alternative Life of Eric Temple Bell**

by Constance Reid

Chreid@aol.com

In the life of E. T. Bell--mathematician, poet, science fiction pioneer, and best-selling author of books about mathematics and mathematicians--there were a dozen important years that he never divulged to any of his Caltech friends and colleagues, nor even to his wife and son--and that he never intended to divulge. As a result the facts of Bell's life (1883-1960) were always erroneously reported until the publication in 1993 of Constance Reid's biographical detective story, *The Search for E.T. Bell, also known as John Taine.*

**Two Remarkable Twos for Inverses to Some Abelian Integrals**

by Peter Lindqvist and Jaak Peetre

lqvist@math.ntnu.no

We refer to some interesting identities generalizing the familiar formula sin^{2 } + cos^{2} = 1 as Ones. Such a formula was proved in 1879 by E. Lundberg for some "sines" and "cosines" arising from the inversion of an Abelian integral. The trigonometric One is a special case. These functions have applications to several branches of analysis.

Our paper is concerned with two related identities called Twos. Our first Two is a generalization of the formula (1 + sl^{2 })(1 + cl^{2 }) = 2 for the lemmiscate functions . The second Two generalizes a formula of A. Cayley from 1882. We also find a curious connection between Cayley's integral and the Weierstrass -function.

**A Chaos Lemma**

by Judy Kennedy, Sahin Kocak, and James a. York

jkennedy@math.udel.edu

The complicated behavior of a trajectory of a dynamical system can often be described in terms of its "itinerary", i.e., the order in which it passes through two (or more) sets, say S_{A} and S_{B}. One trajectory's sequence of A's and B's might be (A,B,B,A,A,A,É), while another's might be quite different. In a chaotic system many different sequences are possible. We begin by describing a simple version of the Smale Horseshoe example. There, the two sets are chosen so that every sequence corresponds to some trajectory. We give a new proof that is more elementary than previous proofs and extends easily to new examples of chaotic systems. The proof is based on families of sets that we call "expanders". It has the flavor of proofs for one-dimensional maps, but in those examples, usually only two expander sets are needed, while in our new proof infinitely many expander sets may be used.

**Real Analyticity: From Calculus Class to the Grauert-Morrey Theorem**

by Robert E. Greene

greene@math.ucla.edu

Because of the uniqueness of analytic continuation, real analytic functions have a kind of "rigidity": a real analytic function on the real line is determined by its values in a neighborhood of any given point, for example. Continuous and even infinitely differentiable functions are much more flexible--they can vanish identically on an open subset of the line without vanishing everywhere. The comparative rigidity of real analytic functions makes many of the standard constructions used in topology very difficult when they are attempted in the real analytic category. This article explains some details of this difficulty and how the difficulty can be overcome by using methods of complex analysis. We begin with a seemingly elementary problem that is easily stated in terms of familiar ideas from calculus. This problem is nonetheless representative of the general questions that arise and of the methods by which they can be treated. Building from this example, we give an outline of how complex analysis methods can be used to prove the Grauert-Morrey Theorem: Every real analytic manifold has a real analytic embedding in some Euclidean space.

**Some Divergent Trigonometric Integrals**

by Erik Talvila

etalvila@math.ualberta.ca

Some years ago a rather famous mathematician made the following error. A convergent integral containing a parameter was differentiated under the integral sign with respect to the parameter without justification. This yielded a divergent integral that is listed even today in standard integral tables as converging. We give a simple proof that the resulting integral diverges and then trace its interesting history.

**NOTES**

**What Goes Up Must Come Down, Eventually**

by Fred Brauer

brauer@math.ubc.ca

**Designing a Calculational Proof of Cantor's Theorem**

by Edsger W. Dijkstra and Jayadev Misra

dijkstra@cs.utexas.edu

**For Every e there Continuously Exists a d**

by Giuseppe De Marco

gdemarco@math.unipd.it

**The Lucas Circles of a Triangle**

by Paul Yiu and Antreas P. Hatzipolakis

yiu@fau.edu

**How to Integrate a Polynomial Over a Sphere**

by Gerald B. Folland

folland@math.washington.edu

**THE EVOLUTION OF ...**

**Foundations of Mathematics in the Twentieth Century**

by V. Wiktor Marek and Jan Mycielski

marek@cs.uky.edu and jmyciel@euclid.colorado.edu

**PROBLEMS AND SOLUTIONS**

**REVIEWS**

**Visual Mathematics, Illustrated by the TI-92 and TI-89.**

By George C. Dorner, Jean Michel Ferrard, and Henri Lemberg

Reviewed by Yves Nievergelt

Yves.Nievergelt@mail.ewu.edu

**Mathematics Success and Failure Among African-American Youth.**

By Danny Bernard Martin

Reviewed by David Scott

scott@ups.edu

**TELEGRAPHIC REVIEWS**