The Line and the Plane

In this chapter we shall examine some topological properties of the real number
line P and the Eucidean plane P2. Some of the ideas presented here will be familiar
from algebra and calculus, but others may be new. In later chapters, the ideas
introduced here will be carried over to a much more general family of sets. Section
2.1 reviews properties of the real number system, and Section 2.2 uses the real
number system to introduce the concept of the number of elements in a set. To-
pological considerations begin formally with consideration of open and closed sets
in Section 2.3.

2.1 UPPER AND LOWER BOUNDS

In order to establish some necessary terminology, let us review the various

types of intervals studied in calculus.
Let a and b be real numbers with a < b. The set
(a,b)= {xEP:aczx<b}
is the open interval with endpoints a and b and

[a,b] =
is the corresponding closed interval. Both of the sets

(a,b]= {xEP:a<xb}
[a,b)= {xEP:ax<b}
are called half-open and half-dosed intervals. The latter two types are distinguished
by the fact that (a, b] is open on the left and closed on the right, while [a, b) is dosed
on the left and open on the right. The sets (a, b), [a, b], (a, b], and [a, b) are called
bounded intervals, since they do not extend indefinitely in either the positive or
negative direction. We shall also have occasion to refer to the unbounded open
intervals

(a, oo) = {xE P: a <x}, a) = {xE P: x <a), (—x, oo) = P

and to the unbounded dosed intervals

[a, oo) = {xE P: a x}, (—ao, a] = {xE P:x a), (—co, ao) = P.
30 TWO / THE LINE AND THE PLANE
Note that intervals from a to b have been defined only for a < b. For the case
of a closed interval, the definition is extended to allow equality of the endpoints
by defining [a, a] to be the singleton set {a}. The empty set 0 is sometimes called
the empty interval. Intervals that are either empty or have only one point are called
degenerate intervals.
The term "interval" is used to refer to sets of the type (a, b), [a, b], (a, b],
[a, b), (a, oo), (—co, a), [a, oo), (—co, a), P, and 0 when no further specialization
in terms of openness, closedness, or boundedness is needed. Open intervals may
be of the bounded type (a, b) and the unbounded types (a, oo), (—oo, a), and P.
Closed intervals may be of the bounded type [a, b] and the unbounded types
[a, oo), (—x), a], and P. A further characterization of intervals will be undertaken
in Chapter 5.

Definition: A number u is an upper bound for a set A of real numbers provided

that a ufor all a E A. If there is a smallest upper bound u0forA, that is, an upper
bound u0 less than all other upper bounds for A, then u0 is called the least upper
bound or supremum of A. The least upper bound for a set A is denoted by lub A or
sup A.

Example 2.1.1
(a) Any real number greater than or equal to 3 is an upper bound for
the set {0, 1, 2, 3}. The least upper bound is 3.
(b) The least upper bound of [a, b] is b.
(c) The least upper bound of(a, b) is b.
(d) The least upper bound of the set (—oo, 0) of negative numbers is 0.
(e) The set(0, co) of positive numbers has no upper bounds and therefore
has no least upper bound.

Note that some sets of real numbers do not have upper bounds and that a
set which has upper bounds may or may not contain its least upper bound. If a set
contains its least upper bound, the least upper bound is simply the largest member
of the set.

Definition: A number 1 is a lower bound for a set A of real numbers provided that
1 afor all a E A. If there is a largest lower bound 10 for A, that is, a lower bound
greater than all other lower bounds for A, then is called the greatest lower bound
or infimum of A. The greatest lower bound of a set A is denoted by glb A or inf A.
 2.1 / Upper and Lower Bounds 31

Example 2.1.2
(a) The greatest lower bound of {O, 1, 2, 3) is 0.
(b) The greatest lower bound of [a, bi is a.
(c) The greatest lower bound of(a, b) is a.
(d) The interval (—ao, 0) has no lower bounds and therefore has no greatest
lower bound.
(e) The greatest lower bound of the interval (0, oo) is 0.

A set may or may not have lower bounds. A set which has lower bounds may
or may not contain the greatest lower bound. If a set contains its greatest lower
bound, then the greatest lower bound is the smallest member of the set.

The Least Upper Bound Property: Every non-empty set of real numbers which
has an upper bound has a least upper bound.

We shall not prove the Least Upper Bound Property but rather accept it as
one of the defining axioms of the real number system. A complete list of the axioms
for the real number system, with corresponding constructive definition of P, can
be found in many textbooks on real analysis. Some references are given in the
suggested reading list at the end of the chapter. It is left as an exercise for the reader
to use this property to prove the corresponding Greatest Lower Bound Property.

The Greatest Lower Bound Property: Every non-empty set of real numbers
which has a lower bound has a greatest lower bound.

Example 2.13
(a) Consider the sequence of real numbers 1, 1/2, 1/3, 1/4.
The least upper bound of this set is 1, the largest member. To see that
0 the greatest lower bound, we reason as follows: Certainly 0 is a
lower bound for the set since 0 < 1/n for each positive integer n.
Now, a number >0 cannot be a lower bound for { because
1/n <€ when n is an integer greater than 1/i. Thus 0 is a lower bound,
and no number greater than 0 is a lower bound. This means that 0
is greater than every other lower bound and is hence the greatest lower
bound of the sequence.
(b) The sequence {1 — of real numbers 0, 1/2, 2/3, 3/4 ... has
greatest lower bound 0 (the smallest term) and least upper bound 1.
32 TWO / THE LINE AND THE PLANE

To see that 1 is the least upper bound, first note that 1 — 1/n < 1 for
every positive integer n, so 1 is an upper bound. It now must be shown
that no number less than 1 is an upper bound. To this end, consider
a number 1 — where 0, and let n be a positive integer greater
than l/E. Then

1 — < 1 — 1/n.

Thus 1 — is not an upper bound for {1 — so 1 must be its

least upper bound.

The next theorem is the first example of an important phenomenon called

denseness, which will appear many times in later chapters. The theorem shows that
there are rational numbers "very close" to every real number. (Recall that a rational
number is a real number which can be expressed as p/q for some integers p, q. A
real number which is not rational is called irrational.)

Theorem 2.1: Between any two real numbers there is a rational number.

Proof: Let a and b be real numbers with a < b. It must be shown that there is a
rational number r with a < r < b. Intuitively, the argument proceeds as follows: Let
q be a positive integer and consider the rational numbers p/q for p = 0, ±1, ±2,
The numbers p/q, when arranged in order on the number line, have successive
terms separated by distance 1/q. If 1/q is less than the distance from a to b, it seems
reasonable that at least one number of the form p/q must fall between a and b.
This intuitive idea is made precise by the upper bound concept. Let q be a
positive integer for which 1/q is less than b — a. The set P = {p/q: p E Z] has no
bounds, either upper or lower. In particular, a is neither an upper bound nor a lower
bound for P. Thus there is an integer Po such that p/q a when p Po andp/q> a
when p >. pj Then the rational number r (Po + 1)/q is greater than a. The fact
that r is less than bfollows from properties of Po and q: p0/q a and 1/q < b — a
so

1
=p0/q+ 1/q<a+(b—a) = b.

Thus r is a rational number between a and b.

Theorem 2.1 can be extended to show that there is an unending sequence of

rational numbers between any two real numbers a and b. There is a rational number
r1 between a and b. There must be, in addition, a rational number r2 between a
and r1 and a rational number r3 between r1 and b. This process can always be
 2.2 / Finite and Infinite Sets 33

repeated to show the existence of another rational number between any two already
determined.
Theorem 2.1 also shows that there are rational numbers within any prescribed
positive distance of a given real number. For a real number a and distance there
must be a rational number between a — and a + Such a rational number will
be at a distance less than from a. The fact that there are rational numbers arbitrarily
close to every real number is expressed by saying that the set of rational numbers
is dense in P.

Example 2.1.4
Let A denote the set of rational numbers between 0 and It should be clear
from the preceding discussion that A has least upper bound and greatest lower
bound 0.

EXERCISE 2.1

1. Find the least upper bound and greatest lower bound, if they exist, for the following sets:
(a) The set Z of integers.
(b) The set of positive integers.
(c) The set of rational numbers greater than
(d) (—1,2)U(3,7).
2. Explain why a set of real numbers cannot have more than one least upper bound or
more than one greatest lower bound.
3. Prove that a set cannot contain more than one of its upper bounds or more than one of
its lower bounds.
4. Prove the Greatest Lower Bound Property assuming the Least Upper Bound Property
as an axiom. (Hint: There is a natural correspondence between upper bounds of a set A
and lower bounds of the set —A of negatives of members of A.)

2.2 FINITE AND INFINITE SETS

Section 2.1 dealt with several subsets of the set of real numbers: integers,
intervals, sequences, rational numbers, and irrational numbers. It was shown that
between any two real numbers there is a rational number. In view of this property,
how many rational numbers are there? Is the number of rational numbers equal