University of Leipzig – WS18/19

10-PHY-BIPMA1 – Mathematics 1 / Vitalii Konarovskyi

4 Lecture 4 – Subsequences and Monotone Sequences

4.1 Monotone Sequences
The main goal of this section is to prove that any bounded monotone sequence must converge. So, we
start from the definition.

Definition 4.1. A sequence (an )n≥1 of real numbers is called an increasing sequence if an ≤ an+1
for all n ≥ 1, and (an )n≥1 is called a decreasing sequence if an ≥ an+1 for all n ≥ 1. A sequence
that is increasing or decreasing is said to be a monotone sequence.

Example 4.1. The sequence (1, 1, 2, 2, 3, 3, 4, 4, . . .) is increasing, but (−1, 1, −1, 1, . . .) is not mono-
tone.

Exercise 4.1. a) Show that any bounded above increasing sequence is bounded. b) Show that any
bounded below decreasing sequence is bounded.

Exercise 4.2. a) Prove that (n2−n )n≥2 is a decreasing sequence.

a1 +...+an
b) Let (an )n≥1 be an increasing sequence of positive numbers and define σn = n . Prove that
(σn )n≥1 is also an increasing sequence.

Theorem 4.1. Every bounded monotone sequence converges.

Proof. We will prove the theorem for increasing sequences. The case of decreasing sequences is left
to Exercise 4.3. So, let a sequence (an )n≥1 increase. By the assumption of the theorem, (an )n≥1 is
bounded, that is, there exists C ∈ R such that |an | ≤ C for all n ≥ 1. This implies that the set
A := {an : n ≥ 1} is also bounded. Thus, by Theorem 2.2 (i) there exists sup A =: sup an denoted by
n≥1
a. Let us prove that an → a, n → ∞. We first note that an ≤ a for all n ≥ 1, since the supremum of A
is also its upper bound (see Definition 2.6). Next, we take an arbitrary ε > 0 and use Theorem 2.1 (i).
So, there exists a number m such that am > a − ε. By the monotonicity, a − ε < am ≤ an for all
n ≥ m. Thus, setting N := m, one has a − ε < an ≤ a for all n ≥ N which implies |an − a| < ε.

Exercise 4.3. Prove Theorem 4.1 for decreasing sequences.

Remark 4.1. Theorem 4.1 remains valid if one requires the monotonicity of (an )n≥1 starting from
some number m, that is, the monotonicity of (an )n≥m = (am , am+1 , . . .).
10n
Example 4.2. Prove that lim = 0, where n! := 1 · 2 · 3 · . . . · n.
n→∞ n!
10 n+1 10n n

Solution. First we note ⇔ 10 < n + 1 ⇔ n > 9. Hence, the sequence 10n! n≥10
that (n+1)! < n!
n

is decreasing. Moreover, it is bounded below by zero. Thus, 10n! n≥10 is bounded, by Exercise 4.1 b).
n
Using Theorem 4.1, one gets that there exists a ∈ R such that lim 10n! = a. But we can write
n→∞
10n+1 10n 10
(n+1)! = n! · n+1 . So,

10n+1 10n 10
a = lim = lim · lim = a · 0.
n→∞ (n + 1)! n→∞ n! n→∞ n + 1

This implies a = 0.

13
University of Leipzig – WS18/19
10-PHY-BIPMA1 – Mathematics 1 / Vitalii Konarovskyi

n! n
Exercise 4.4. Show that a) lim 2 = 0; b) lim √
n = 0.
n→∞ 2n n→∞ 2

√ p √ √
 q p 
Exercise 4.5. Find a limit of the sequence 2, 2 + 2, 2 + 2 + 2, . . . .

Exercise 4.6. Let a1 = 1 and an+1 = 13 (an + 1) for all n ≥ 1.

a) Find a2 , a3 , a4 .
1
b) Use induction to show that an > 2 for all n ≥ 1.

c) Show that (an )n≥1 is a decreasing sequence.

d) Show that lim an exists and find it.

n→∞
 
1 c
Exercise 4.7. Let c > 0, a1 > 0 and let an+1 = 2 an + an for all n ≥ 1.
√
a) Show that an ≥ c for all n ≥ 2.

b) Show that (an )n≥2 is a decreasing sequence.

c) Show that lim an exists and find it.

n→∞

Theorem 4.2. (i) If (an )n≥1 is an unbounded increasing sequence, then lim an = +∞.
n→∞

(ii) If (an )n≥1 is an unbounded decreasing sequence, then lim an = −∞.

n→∞

Proof. We will prove only Part (i) of the theorem. The proof of Part (ii) is similar. If (an )n≥1 is
an unbounded increasing sequence, then it must be unbounded above, since it is bounded below by
a1 . Taking any C and using the unboundedness of (an )n≥1 , one can find a number m ∈ N such that
am ≥ C. Next, by the monotonicity of (an )n≥1 , the inequality an ≥ am ≥ C trivially holds for all
n ≥ N := m. This proves lim an = +∞ (see Definition 3.4).
n→∞

Corollary 4.1. If (an )n≥1 is a monotone sequence, then the sequence either converges, diverges to
+∞, or diverges to −∞. Thus lim an is always meaningful for monotone sequences.
n→∞

Proof. The proof immediately follows from theorems 4.1 and 4.2.

Exercise 4.8. Let A be a bounded nonempty subset of R such that sup A is not in A. Prove that
there is an increasing sequence (an )n≥1 of points from A such that lim an = sup A.
n→∞

4.2 The number e

In this section, we will consider two sequences of positive numbers
  n   n+1 !
1 1
an := 1+ and bn := 1+ (1)
n n≥1 n
n≥1

and study their properties.

14
University of Leipzig – WS18/19
10-PHY-BIPMA1 – Mathematics 1 / Vitalii Konarovskyi

Theorem 4.3. The sequences defined in (1) satisfy the following properties:
1) an < bn for all n ≥ 1;
2) the sequence (an )n≥1 increases;
3) the sequence (bn )n≥1 decreases.
Proof. Since bn = an 1 + n1 = an + ann > an for all n ≥ 1, Property 1) is proved.

To prove 2), we are going to use Bernoulli’s inequality (see Theorem 2.6). So, one has
n + 1 n n − 1 n−1 1 n
     
an n n  n
= = 1− 2 > 1 − 2 = 1,
an−1 n n n−1 n n−1 n
for all n ≥ 2. Thus, an > an−1 for all n ≥ 2.
For the prove of 3) we use the same argument. We consider
n  n+1 n+1
n2
 
bn−1 n n n−1
= =
bn n−1 n+1 n n2 − 1
 n+1  
n−1 1 n−1 n+1
= 1+ 2 > 1+ 2 = 1,
n n −1 n n −1
for all n ≥ 2. Hence, bn−1 > bn for all n ≥ 2.

Theorem 4.3 yields the following inequalities

a1 < a2 < . . . < an < . . . < bn < . . . < b2 < b1 . (2)
Consequently, the sequences (an )n≥1 and (bn )n≥1 are monotone and bounded. By Theorem 4.1, they
converge. We set
1 n
 
e := lim an = lim 1 + = 2, 718281828459045...
n→∞ n→∞ n
It is known that e is an irrational number. The number e is one of the most important constants in
mathematics.
Since bn = an 1 + n1 for all n ≥ 1, one has bn → e, n → ∞. We also note that

1 n 1 n+1
   
1+ <e< 1+ , (3)
n n
by inequalities (2).
Definition 4.2. The logarithm to base e is called the natural logarithm and is denoted by ln := loge ,
that is, for each a > 0 ln a is a (unique!) real number such that eln a = a.
The inequality (3) immediately implies
 
1 1 1
< ln 1 + <
n+1 n n
for all n ≥ 1.
1



Exercise 4.9. Show that lim n ln 1 + n = 1.
n→∞
x n



Exercise 4.10. Prove that for each x > 0 the sequence 1+ n n≥1
is increasing and bounded.

15
University of Leipzig – WS18/19
10-PHY-BIPMA1 – Mathematics 1 / Vitalii Konarovskyi

4.3 Subsequences
4.3.1 Subsequences and Subsequential Limits
Let (an )n≥1 be a sequence. We consider any subsequence (nk )k≥1 of natural numbers such that
1 ≤ n1 < n2 < . . . < nk < nk+1 < . . .. We note that nk ≥ k and nk → +∞, k → ∞.

Example 4.3. 1) nk = k, k ≥ 1; then (nk )k≥1 = (1, 2, 3, . . . , k, . . .);

2) nk = 2k, k ≥ 1; then (nk )k≥1 = (2, 4, 6, . . . , 2k, . . .);

3) nk = k 2 , k ≥ 1; then (nk )k≥1 = (1, 2, 9, . . . , k 2 , . . .);

4) nk = 2k , k ≥ 1; then (nk )k≥1 = (2, 4, 8, . . . , 2k , . . .).

Definition 4.3. A sequence (ank )k≥1 = (an1 , an2 , an3 , . . . , ank , . . .) is said to be a subsequence of
(an )n≥1 .

Thus, (ank )k≥1 is just a selection of some (possibly all) of the an ’s taken in order.

Remark 4.2. The following properties follows from the definition of subsequence.

1. If a sequence is bounded, then every its subsequence is bounded.

2. If a sequence converges to a (that could be +∞ or −∞), then every its subsequence also converges
to a.

Exercise 4.11. Prove that a monotone sequences which contains a bounded subsequence is bounded.

Exercise 4.12. Prove that a sequence (an )n≥1 converges iff (a2k )k≥1 , (a2k−1 )k≥1 and (a3k )k≥1 con-
verge.

Definition 4.4. A subsequential limit of a sequence (an )n≥1 is any real number or the symbol +∞
or −∞ that is the limit of some subsequence of (an )n≥1 . Let A denotes the set of all subsequential
limit of (an )n≥1 .

Example 4.4. a) For the sequence (1, 2, 3, . . . , n, . . .) the set of all subsequential limit A = {+∞}.

b) For the sequence (−1, 1, −1, . . . , (−1)n , . . .) the set of all subsequential limit A = {−1, 1}.

c) If an → a, then A = {a}, by Remark 4.2.

Exercise 4.13. Prove the following statements.

a) −∞ ∈ A ⇔ (an )n≥1 is unbounded below. b) +∞ ∈ A ⇔ (an )n≥1 is unbounded above.

Exercise 4.14. Find the set A of all subsequential limits of the following sequences.
n+1
(
(−1) 2 + n, if n is odd,
a) (sin 3πn)n≥1 ; b) (sin απn)n≥1 for α ∈ Q; c) (an )n≥1 , where an = n
(−1) 2 + n1 , if n is even.

16
University of Leipzig – WS18/19
10-PHY-BIPMA1 – Mathematics 1 / Vitalii Konarovskyi

4.3.2 Existence of Monotone Subsequence

Theorem 4.4. A number a ∈ R is a subsequential limit of a sequence (an )n≥1 iff

∀ε > 0 ∀N ∈ N ∃ñ ∈ N : ñ ≥ N, |añ − a| < ε. (4)

Proof. We first prove the necessity. Let a ∈ A. Then there exists a subsequence (ank )k≥1 such that
ank → a, k → ∞. We fix an arbitrary ε > 0 and N ∈ N. By the definition of the limit, ∃K1 ∈ N ∀k ≥
K1 : |ank − a| < ε. Similarly, ∃K2 ∈ N ∀k ≥ K2 : nk ≥ N. Thus, taking k̃ := max{K1 , K2 }, ñ := nk̃ ,
one has ñ ≥ N and |añ − a| < ε.
To prove the sufficiency, we are going to construct a subsequence of (an )n≥1 converging to a. Let (4)
holds. Then, by (4), for ε = 1 and N = 1 there exists n1 ≥ 1 such that |an1 − a| < 1. Similarly, for
ε = 21 and N = n1 + 1 there exists n2 ≥ n1 + 1 such that |an2 − a| < 21 and so on. Consequently,
we obtain a subsequence (ank )k≥1 satisfying |ank − a| < k1 for all k ≥ 1. Using Theorem 3.7 and
Exercise 3.5 a), one can see that ank → a, k → ∞.

Exercise 4.15. Show that +∞ ∈ A (−∞ ∈ A) provided ∀C ∈ R ∀N ∈ N ∃ñ ∈ N : ñ ≥ N and

añ ≥ C (añ ≤ C).

Theorem 4.5. Every sequence of real numbers contains a monotone subsequence.

Proof. We consider the set M := {n ∈ N : ∀m > n am > an }. If M is infinite, then M can be

written as M = {n1 , n2 , . . . , nk , . . .}, where n1 < n2 < . . . < nk < . . .. By the definition of M , we have
an1 < an2 < . . . < ank < . . .. So, the subsequence (ank )k≥1 increases.
If M is finite, then let n1 be the smallest natural number such that ∀m ≥ n1 : m 6∈ M . Since
n1 6∈ M , one can find n2 > n1 such that an1 ≥ an2 . Similarly, since n2 6∈ M , one can find n3 > n2
such that an2 ≥ an3 and so on. Thus, the constructed subsequence (ank )k≥1 decreases.

Corollary 4.2. For every sequence the set of its subsequential limits is not empty.

Proof. The corollary immediately follows from Theorem 4.5 and Corollary 4.1.

Theorem 4.6 (Bolzano-Weierstrass theorem). Every bounded sequence has a convergent subsequence.

Proof. The theorem is a direct consequence of theorems 4.5 and 4.1.

References
[1] K.A. Ross. Elementary Analysis: The Theory of Calculus. Undergraduate Texts in Mathematics.
Springer New York, 2013.

17