Part 1, Semester 3.

Series of Real Numbers *

Leonid Pankratov

July 4, 2023

It is impossible to be a mathematician without being a poet in soul.

Sofia Kovalevskaya, Russian mathematician.

* File :SER–REAL–NUM.tex

1
 PROGRAM of the COURSE.

2. Numerical series. Cauchy criterion for convergence of a numerical series. Series of non-negative
terms: comparison theorems, Cauchy’s test, d’Alembert’s test, the integral test for convergence of a
series. Absolute and conditional convergence, Abel’s and Dirichlet’s tests for convergence of a series.
Reordering of series’ terms: independence of the sum for absolutely convergent series, Riemann
series theorem.

2
Contents
1 Basic Definitions 4

2 Convergence and Divergence of Series 4

3 Convergence Tests 6

4 Alternating Series 12
4.1 Abel’s transform formula on summation by parts . . . . . . . . . . . . . . . . . . . . . . . . 13
4.2 Dirichlet and Abel convergence tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

5 Absolute Convergence 16

6 Rearrangement of series 17

7 The Series With Complex Terms 18

8 Solved Problems 19

3
1 Basic Definitions

Definition 1.1 (Series of Real Numbers) A series of real numbers is an expression of the form

a1 + a2 + a3 + ...

or more compactly
∞
X
an ,
n=1
where {an } is a sequence of real numbers.

Definition 1.2 (n-th Partial Sum) The number an is called the n-th term of the series and the se-

quence Sn ,
n
X
def
Sn = ak
k=1
∞
X
is called the n-th partial sum of the series an .
n=1

2 Convergence and Divergence of Series

∞
X
Definition 2.1 (Convergence and Divergence of a Series) A series an is said to converge to
n=1
S ∈ R if the sequence {Sn } of partial sums of the series converge to S ∈ R.
X ∞
If an converges to S, then we will write
n=1
∞
X
an = S.
n=1

A series which does not converge is called a divergent series.

Now we formulate the necessary condition of convergence.
∞
X
Theorem 2.2 (Necessary Condition of Convergence) If a series an converge, then
n=1

an → 0 as n → ∞.
∞
X
Proof of Theorem 2.2. Clearly, if Sn is the n-th partial sum of the convergent series an , then
n=1

an = Sn − Sn−1 → 0 as n → ∞.
Theorem 2.2 is proved.

4
Remark 1 To see that the converse does not hold (that is this condition is not sufficient) it is enough
to observe that the series
∞
X 1
an with an = (∀n ∈ N)
n=1
2n − 1
diverges whereas an → 0 as n → ∞.
In fact, the sequence of partial sums of this series is monotonically increasing. Let us show that

this sequence is not bounded. To this end we consider its subsequence S2n , n ∈ N, we have :
1 1 1 1 1 1
S21 = S2 = 1 + ; S22 = S4 = 1 + + + ; ...; S2n = 1 + + ... + n+1 .
3 3 5 7 3 2 −1
Now it follows from the estimates :
1 1 1 2 1 1 1 1 2n−1 1
1+ > 1; + > = ; ...; n
+ n
+ ... + n+1
> n+1
= .
3 5 7 8 4 2 +1 2 +3 2 −1 2 4
Then
     
1 1 1 1 1 1 n−1
S2n = 1+ + + + ... + n
+ n + ... + n+1 . >1+
3 5 7 2 +1 2 +3 2 −14

This means that the subsequence {S2n } is not bounded. This implies that the sequence Sn is also
not bounded. Thus the series is divergent.

Example 2.3 The harmonic series,

∞
X 1
n=1
n
is divergent.
In fact :
2n
X 1 1 1
S2n − Sn = ⩾ · n = ̸→ 0 as n → ∞.
k=n+1
k 2n 2
 
This contradicts to the fact that in the case of convergent series, the sequences Sn , S2n have to
converge to the same limit and their difference to zero.
The proof of the following corollary is immediate from the above theorem.

Corollary 2.4 Suppose an is a sequence of positive terms such that

an+1 > an ∀n ∈ N.
∞
X
Then the series an diverge.
n=1

5
  
Theorem 2.5 Let an and bn be sequences such that there is K ∈ N,

an = b n ∀n ⩾ K.
∞
X ∞
X
Then an converges if and only if bn is convergent.
n=1 n=1

∞
X ∞
X
Proof of Theorem 2.5. Suppose An and Bn be n-th partial sums of the series an , bn ,
n=1 n=1
respectively. Let
K
X K
X
def def
α= ai and β = bi .
n=1 n=1

Then we have : n n
X X
An − α = ai = bi = Bn − β ∀n ⩾ K.
i=K+1 i=K+1
 
It follows from this relation that the sequence An converges if and only if Bn converges.
Theorem 2.5 is proved.
∞
X
P∞
The above theorem implies that if b
n=1 n is obtained from an
n=1

∞
X ∞
X
an converges ⇔ bn converges.
n=1 n=1

The proof of the following theorem is evident.

∞
X ∞
X
Theorem 2.6 Suppose an converges to A and bn converges to B. Then for any α, β ∈ R,
n=1 n=1

∞ 
X 

αan + βbn converges to αA + βB .
n=1

3 Convergence Tests
 
Theorem 3.1 (Comparison Test) Suppose an and bn are sequences with non-negative terms,
and an ⩽ bn for all n ∈ N. Then,
∞
X ∞
X
(1) If the series bn converges =⇒ an converges.
n=1 n=1

∞
X ∞
X
(2) If the series an diverges =⇒ bn diverges.
n=1 n=1

6
 ∞
X
Proof of Theorem 3.1. Suppose An and Bn are the n-th partial sums of the series an and
n=1
∞
X
bn , respectively. By the assumption, we get
n=1

0 ⩽ An ⩽ Bn ∀n ∈ N
and both An , Bn are monotonically increasing.
Statement (1). Since the sequence Bn is convergent, it is bounded. Let M > 0 be such that Bn ⩽ M
for all n ∈ N. Then An ⩽ M for all n ∈ N. Since An is monotonically increasing, it follows that An
converges.
Statement (2). The proof of the statement is evident and can be done by contradiction. In fact, if the
∞
X ∞
X
series bn converges, then by the first part of the theorem the series an is also convergent.
n=1 n=1
Theorem 3.1 is proved.

 
Corollary 3.2 Suppose an and bn are sequences with positive terms and
def an
ℓ = lim (3.1)
n→∞ bn

exists. Then
∞
X ∞
X
1. If ℓ > 0, then bn converges ⇐⇒ an converges.
n=1 n=1
∞
X ∞
X
2. If ℓ = 0, then if bn converges =⇒ an converges.
n=1 n=1

Proof of Corollary 3.2. We suppose that (3.1) is fulfilled and ℓ > 0. Then for any ε > 0 there exists
N ∈ N such that
an
ℓ−ε< < ℓ + ε ∀n ⩾ N
bn
or
(ℓ − ε)bn < an < (ℓ + ε)bn ∀n ⩾ N.
Now we set ε = ℓ/2 to obtain that
ℓ 3ℓ
bn < an < bn ∀n ⩾ N.
2 2
Hence, the result follows by comparison test.
Let ℓ = 0. Then for any ε > 0 there exists N ∈ N such that
an
−ε < < ε ∀n ⩾ N.
bn
In particular,
an < εbn ∀n ⩾ N.
Hence, we get the result by using comparison test. Corollary 3.2 is proved.

7
 
Theorem 3.3 (D’Alembert Ratio Test) Suppose that an is a sequence of positive real numbers
such that
an+1
lim =ℓ
n→∞ an

exists. Then
∞
X
(1) If ℓ < 1, then an converges.
n=1

∞
X
(2) If ℓ > 1, then an diverges.
n=1

Proof of Theorem 3.3. Consider statement (1). Suppose ℓ < q < 1. Then there exists N ∈ N such
that
an+1
< q ∀n ⩾ N.
an
In particular,
an+1 < qan < q 2 an−1 < ... < q n a1 ∀n ⩾ N.
∞
X ∞
X
n
Since q converges, by comparison test, an is also convergent.
n=1 n=1

Consider statement (2). Let

1 < p < ℓ.
Then there exists N ∈ N such that
an+1
>p>1 ∀n ⩾ N.
an
∞
X
This inequality implies that an does not converge to zero (an+1 > an > 0). Hence an diverges.
n=1
Theorem 3.3 is proved.

Theorem 3.4 (Cauchy’s Root Test) Suppose {an } is a sequence of positive terms such that

lim (an )1/n = ℓ

n→∞

exists. Then
∞
X
(1) If ℓ < 1, then an converges.
n=1

∞
X
(2) If ℓ > 1, then an diverges.
n=1

8
Proof of Theorem 3.4. Consider statement (1). Suppose ℓ < q < 1. Then there exists N ∈ N such
that
(an )1/n < q ∀n ⩾ N.
Hence
an < q n ∀n ⩾ N.
∞
X ∞
X
n
Since q converges, by comparison test, the series an also converges.
n=1 n=1

Consider statement (2). Let 1 < p < ℓ. Then there exists N ∈ N such that

(an )1/n > p > 1 ∀n ⩾ N.

∞
X
Hence an ⩾ 1 for all n ⩾ N . Thus an does not converge to zero. Hence an diverges. Theorem
n=1
3.4 is proved.

Remark 2 We remark that both d’Alembert’s and Cauchy’s tests are silent for the case ℓ = 1. But,
for such case, we may be able to infer the convergence or divergence by some other means.

Example 3.5 Consider the series

∞
X xn
,
n=1
n!
where x is a fixed real number. Let us show that this series converges. To this end, we make use of the
d’Alembert’s test. In this example
def xn
an = .
n!
Then
an+1 xn+1 n! x
= · n =
an (n + 1)! x n+1
Hence we have that
an+1
lim =0
n→∞ an

and by d’Alembert’s test we conclude that the series is convergent.

Example 3.6 Show that the series

∞  n
X n
n=1
2n + 1
converges.

9
 In fact, here
n 1
a1/n
n = −→ < 1.
2n + 1 2
Hence, by Cauchy’s test, the series converges.

Theorem 3.7 (Integral test for convergence) If f is a continuous, positive and decreasing function
where f (n) = an on the interval [1, ∞), then the improper integral

Z+∞ ∞
X
f (x) dx and the infinite series an
1 n=1

either both converge or both diverge.

Proof of Theorem 3.7. Picture infinitely many rectangles of width 1 and height an , so the area of the
∞
X
n-th rectangle is an . Then the series an is equal to the sum of the areas of these infinitely many
n=1
rectangles. See the graphic examples below.

Figure 1: Integral test. Figure 1

10
 Figure 2: Integral test. Figure 2

Consider Fig. 1. We see that the value of the series from a2 on is less than the area under the curve
f from 1 to infinity, i.e.,
X∞ Z∞
an < f (x) dx
n=2 1

If this integral converges to some finite value C, then by using this inequality, we get
∞
X ∞
X Z∞
an = a1 + an < a1 + f (x) dx = a1 + C < ∞,
n=1 n=2 1

so the series is finite, and thus it converges.

Consider Fig. 2. We see that the sum of the same series beginning with a1 is larger than the area
under the same curve f from 1 to infinity, i.e.,
Z∞ ∞
X
f (x) dx < an
1 n=1

If this integral diverges, then because of our constraints on f it diverges to infinity. Since P the area
under f is infinite, then the sum of the areas of the rectangles must also be infinite, i.e, ∞ n=1 an is
infinite, and thus the series diverges. We see that if the integral diverges, so does the series.
The conclusion is : either both the integral and the series converge, or both diverge.

11
4 Alternating Series

Definition 4.1 A series of the form

∞
X
(−1)n+1 an
n=1

where {an } is a sequence of positive real numbers is called an alternating series.

Theorem 4.2 (Leibniz’s Theorem) Suppose {an } is a sequence of positive terms such that

an ⩾ an+1 (4.1)

for all n ∈ N, and

an → 0 as n → ∞.
∞
X
Then the alternating series (−1)n+1 an converges.
n=1

∞
X
Proof of Theorem 4.2. Let Sn be the n-th partial sum of the alternating series (−1)n+1 an . We
n=1
observe that
S2n+1 = S2n + a2n+1∀n ∈ N.

Since an → 0 as n → ∞, it is enough to show that the sequence S2n converges. Notice that

S2n = (a1 − a2 ) + (a3 − a4 ) + ... + (a2n−1 − a2n ). (4.2)

Then from (4.1) it follows that S2n > 0 and the sequence is increasing. From the other hand, we can
rewrite S2n as follows :

S2n = a1 − (a2 − a3 ) − ... − (a2n−2 − a2n−1 ) − a2n

for all n ∈ N. Here any term is positive, then

S2n < a1 .

Thus the sequence S2n is increasing and bounded above. Thus this sequence is convergent. We
can write
S2n → S as n → ∞.
Then since an → 0 as n → ∞, we have that

S2n+1 = S2n + a2n+1 → S as n → ∞.

 
Now it follows from the convergence of S2n and S2n+1 to the same limit that

Sn → S as n → ∞.

This completes the proof of Theorem 4.2.

12
Example 4.3 Theorem 4.2 implies that the series
∞ ∞ ∞
X (−1)n+1 X (−1)n+1 X (−1)n+1
, , and
n=1
n n=1
2n − 1 n=1
2n

are also convergent.

Now we will establish convergence tests for the series of the form :
∞
X
ak b k . (4.3)
k=1

First, we obtain an auxiliary result.

4.1 Abel’s transform formula on summation by parts

We consider the transform of the finite sum (Abel’s transform):
n
X
Sn = ak b k .
k=1

Let us introduce the notation :

k
X
def
B0 ∈ R, Bk = B0 + bj (k = 1, 2, ..., n). (4.4)
j=1

Then
bk = Bk − Bk−1 (k = 1, 2, ..., n)
and n n n n
X X   X X
Sn = ak b k = ak Bk − Bk−1 = ak Bk − ak Bk−1 . (4.5)
k=1 k=1 k=1 k=1

Consider the second term on the right-hand side of the relation (4.5). We change the index k to have :

k − 1 := ℓ;
 
n
X n−1 n−1 h i
 X X
ak Bk−1  k = 1 ⇒ ℓ = 0; = aℓ+1 Bℓ = a1 B0 + aℓ+1 Bℓ ℓ := k =

k=1 ℓ=0 ℓ=1
k =n⇒ℓ=n−1
n−1
X
= a1 B0 + ak+1 Bk . (4.6)
k=1

Now from (4.5), (4.6) we obtain that

n
X n
X n−1
X n−1
X
Sn = ak Bk − ak Bk−1 = an Bn + ak Bk − a1 B0 − ak+1 Bk =
k=1 k=1 k=1 k=1

13
 n−1 
X 
= an Bn − a1 B0 − ak+1 − ak Bk .
k=1

Thus
n
X n−1 
X 
Sn = ak bk = an Bn − a1 B0 − ak+1 − ak Bk . (4.7)
k=1 k=1

Formula (4.7) is known as Abel’s transform formula on summation by parts.

Now we will consider the well-known tests for the series (4.3).

4.2 Dirichlet and Abel convergence tests


Theorem 4.4 (Dirichlet Test) Suppose the sequence ak goes to zero monotonically and the se-
∞
X ∞
X
quence of n-th partial sums of bk is bounded. Then the series ak bk converges.
k=1 k=1

Proof of Theorem 4.4. Using notation (4.4) with B0 = 0, let us apply Abel’s transform (4.7) to the
∞
X
partial sum of ak b k .
k=1
We study the behavior of the right-hand side of (4.7). First, we observe that due to the boundedness
of the sequence {Bn } and the condition an → 0 as n → ∞,

an Bn → 0 as n → ∞.

The sum on the right-hand side of (4.7) has a finite limit. In fact, setting
def
M = sup |Bk |,
k∈N

one obtains that

∞ 
X  ∞ 
X  ∞ 
X 
ak+1 − ak Bk = M ak+1 − ak =M ak+1 − ak = M |a1 |.
k=1 k=1 k=1

∞
X
Therefore, the left-hand side of (4.7) converges to the same limit and the series ak bk converges.
k=1
Theorem 4.4 is proved.

Corollary 4.5 In the framework of Theorem 4.4,

∞
X n
X
ak bk ⩽ |a1 | sup bk .
n∈N
k=1 k=1

14
 
Theorem 4.6 (Abel Test) Suppose the sequence ak is monotonic and bounded and the series
X∞ ∞
X
bk converges. Then the series ak bk converges.
k=1 k=1

Proof of Theorem 4.4. Suppose

ak → a0 as k → ∞.
Then
def
α k = ak − a0 → 0 as k → ∞
and ∞ ∞ ∞
X X X
ak b k = α k b k + a0 bk .
k=1 k=1 k=1

The first series on the right-hand side of this relation converges by the Dirichlet test above and the
X∞
second one by the conditions of our theorem. This implies that the series ak bk converges.
k=1
Theorem 4.4 is proved.

Example 4.7 Let us show that the series

∞
X sin kx
with α > 0, x ∈ R (4.8)
k=1
kα

is convergent.
The proof will be done with the help of Dirichlet test (Theorem 4.4). First, we observe that the
sequence
def 1
ak = (α > 0)
kα
monotonically goes to zero. Show that the sequence of partial sums
n
X
sin kx
k=1

is bounded. In fact, for x ̸= 2πm (m ∈ Z),

n n n      
X 1 X x 1 X 1 1
sin kx = 2 sin · sin kx = cos k − x − cos k + x .
k=1
2 sin x2 k=1 2 2 sin x2 k=1 2 2

This relation implies that

n
X 1
sin kx ⩽ .
k=1
sin x2
Thus, the convergence of series (4.8) follows from the Dirichlet test if x ̸= 2πm (m ∈ Z). If
x = 2πm (m ∈ Z) the series (4.8) converges because all the terms of this series are zeros.

15
 In a similar way, one can show the convergence of
∞
X cos kx
with α > 0, x ∈ R. (4.9)
k=1
kα

Example 4.8 Let us show that the series

∞
X sin kx
ak with α > 0, x ∈ R,
k=1
kα

where {ak } is a monotonic bounded sequence, converges.

Clearly, this series converges by Abel’s test.

5 Absolute Convergence
∞
X ∞
X
Definition 5.1 A series ak is said to converge absolutely, if |ak | converges.
k=1 k=1

Theorem 5.2 Every absolutely convergent series converges.

∞
X
Proof of Theorem 5.2. Suppose ak is an absolutely convergent series. Let sn and σn be the n-th
k=1
∞
X ∞
X
partial sums of the series ak and |ak |, respectively. Then, for n > m, we have
k=1 k=1

n
X n
X
sn − sm = an ⩽ |an | = σn − σm .
j=m+1 j=m+1

 
Since, σn converges, it is a Cauchy sequence. Hence, form the above relation it follows that sn
is also a Cauchy sequence. Therefore, by the Cauchy criterion, it converges. Theorem 5.2 is proved.

∞
X ∞
X
Definition 5.3 A series ak is said to converge conditionally if ak converges but not abso-
k=1 k=1
lutely.

16
6 Rearrangement of series
Notice that a finite sum is not changed by rearranging its terms, i.e.,

1 + 2 + 3 = 1 + 3 + 2 = 2 + 3 + 1 = ... = 6.

This is not true of all infinite series.

X X
Definition 6.1 Let us say that bn is a rearrangement of an if the two series have the same
terms, written in possibly different orders.

Notice that since the partial sums of the two series may form entirely different sequences, there is
no reason to expect them to exhibit the same convergence properties, and in general it is not the fact.
We are interested in what happens if we rearrange the terms of a convergent series. We will see
that every rearrangement of an absolutely convergent series has the same sum, but that conditionally
convergent series fail, spectacularly, to have this property.
∞
X ∞
X ∞
X
Theorem 6.2 If bn is a rearrangement of an absolutely convergent series an , then bn
n=1 n=1 n=1
is also converges absolutely and, moreover, to the same sum.

Proof of Theorem 6.2. Let us introduce the notation :

def def
A
en = a1 + a2 + ... + an and B
en = b1 + b2 + ... + bn .

For each n ⩾ 1, there is an integer kn such that b1 , b2 , ..., bn are included among a1 , a2 , ..., akn .
Therefore,
B
en ⩽ Ae kn .
∞
X
 
Since An is bounded, then Bn is also bounded and, therefore,
e e bn < ∞.
n=1
Now let
def def
An = a1 + a2 + ... + an and Bn = b1 + b2 + ... + bn ;
∞
X ∞
X
A= an and B = bn .
n=1 n=1

We must show that A = B. Suppose that ε > 0. From Cauchy’s convergence criterion for series
∞
X
and the absolute convergence of an , there exists an integer N = N (ε) such that
n=1

aN +1 + aN +2 + ... + aN +k < ε, k ⩾ 1.

Choose N1 so that the terms a1 , a2 , ..., aN are included among b1 , b2 , ..., bN1 . If n ⩾ N1 , then An and
Bn both include the terms a1 , a2 , ..., aN , which cancel on subtraction. Thus, An − Bn is dominated

17
 ∞
X
by the sum of the absolute values of finitely many terms from an with subscripts greater than N .
n=1
Since every such sum is less than ε,

An − Bn < ε if n ⩾ N1 .

Therefore,

lim An − Bn = 0 and A = B.
n→∞

This completes the proof of Theorem 6.2.

∞
X
Theorem 6.3 (Riemann rearrangement theorem) Suppose that the series an is convergent
n=1
but not absolutely. Then for any A ∈ R, it is possible to rearrange the terms in such a way that the
obtained series converges to A.

7 The Series With Complex Terms

Definition 7.1 The sequence of complex numbers

 
zk = xk + iyk

is said to be convergent if there exists a complex number z0 = x0 + iy0 such that

lim zk − z0 = 0.
k→∞


The complex number z0 = x0 + iy0 is called the limit of sequence zk .

Definition 7.2 The symbol

∞
X
zk zk ∈ C
k=1

is called the series with complex terms.

18
8 Solved Problems

Exercise 8.1 Determine whether

∞
X √
n
3
n=1

converges or diverges.

Solution to Exercise 8.1. In this exercise we make use of the n-th term test :
∞
X
If the series an is convergent, then lim an = 0.
n→∞
n=1
It is important to notice that the converse of this theorem is false. The convergence of an to zero
X ∞
does not imply that the series an converges.
n=1
Since √
n 1
lim 3 = lim 3 n = 1,
n→∞ n→∞

then the series ∞

X √
n
3
n=1

diverges by the n-th term test.

Exercise 8.2 Study the convergence of series

∞
X en
.
n=1
n2

Solution to Exercise 8.2. Calculate the limit

en
lim .
n→∞ n2

Using L’Hopital’s rule, we find

ex ex ex
lim = lim = lim = ∞.
x→∞ x2 x→∞ 2x x→∞ 2

Hence, the original series diverges by the n-th term test.

Exercise 8.3 Study the convergence of series

∞  
X 1 1
+ .
n=0
3n 5n

19
Solution to Exercise 8.3. This series converges because it is the sum of two convergent series,
∞ ∞
X 1 X 1
and
n=0
3n n=0
5n

Both are geometric series with ratio |q| < 1. The general formula in this case, when the zero term of
the progression is one, is given by
∞
X 1
qn = . (8.1)
n=0
1−q
Then it follows from (8.1) that
∞ ∞  n h
X 1 X 1 1i 1 3
= q = = 1 =
n=0
3n n=0
3 3 1− 3
2

and ∞ ∞  n h
X 1 X 1 1i 1 5
= q = = 1 = .
n=0
5n n=0
5 5 1− 5
4
Hence, the sum of the given series is
∞  
X 1 1 3 5 11
n
+ n = + = .
n=0
3 5 2 4 4

Exercise 8.4 Study convergence of the series

∞
X 1
.
n=1
(n + π) (n + π + 1)

Solution to Exercise 8.4. We see that

1 1 1
= − .
(n + π) (n + π + 1) n+π n+π+1

Then the n-th partial sum is

     
1 1 1 1 1 1 1 1
Sn = − + − + . . .+ − = − .
1+π 2+π 2+π 3+π n+π n+π+1 1+π n+π+1

Calculate the limit of Sn as n → ∞ :

 
1 1 n 1
lim Sn = lim − = lim = .
n→∞ n→∞ 1+π n+π+1 n→∞ (1 + π)(n + π + 1) 1+π

Hence, the series converges.

20
Exercise 8.5 Determine whether the series

1 1 1 1 1
+ + + + ... + + ...
1·2 2·3 3·4 4·5 n (n + 1)

converges or diverges.

Solution to Exercise 8.5. The n-th partial sum is

1 1 1 1 1
Sn = + + + + ... + .
1·2 2·3 3·4 4·5 n (n + 1)
We can easily see that
1 1 1 1 1 1 1 1 1 1 1 1 1 1
=1− ; = − ; = − ; = − ; ...; = − .
1·2 2 2·3 2 3 3·4 3 4 4·5 4 5 n (n + 1) n n+1
Then        
1 1 1 1 1 1 1
Sn = 1 − + − + − + ... + − .
2 2 3 3 4 n n+1
Hence,  
1 1
Sn = 1 − and lim Sn = lim 1 − = 1.
n+1 n→∞ n→∞ n+1
Thus the series converges to 1.

Exercise 8.6 Determine whether the series

∞
X n+2
ln .
n=0
n+1

converges or diverges.

Solution to Exercise 8.6. We write the n-th term as

n+2
an = ln = ln (n + 2) − ln (n + 1) .
n+1
Calculate the n-th partial sum:

Sn = (ln 2 − ln 1) + (ln 3 − ln 2) + (ln 4 − ln 3) + . . . + [ln (n + 2) − ln (n + 1)] =

= (− ln 1 + ln 2) + (− ln 2 + ln 3) + (− ln 3 + ln 4) + . . . + [− ln (n + 1) + ln (n + 2)] =
= − ln 1 + ln (n + 2) = ln (n + 2) .
Since
lim Sn = lim ln (n + 2) = +∞,
n→∞ n→∞

we conclude that the given series diverges.

21
Exercise 8.7 Determine whether the series

def
U = q sin α + q 2 sin 2α + ... + q n sin nα + ...

and
def
V = q cos α + q 2 cos 2α + ... + q n cos nα + ...

where |q| < 1 converge or diverge and find their sums.

Solution to Exercise 8.7. Denote by Un and Vn the partial sums of the series U and V , respectively.
Then taking into account that
def
√

eiφ = cos φ + i sin φ i = −1 ,

we get
qeiα − q n+1 ei(n+1)α
Un + iVn = qeiα + q 2 ei2α + ... + q n einα = . (8.2)
1 − qeiα
Taking into account that |q| < 1, we have that |qeiα | < 1. Then we have that
 
lim q n+1 ei(n+1)α = 0.
n→∞

Now formula (8.2) implies that

qeiα
 
  cos α − q sin α
U + iV = lim Un + iVn = =q +i .
n→∞ 1 − qeiα 1 − 2q cos α + q 2 1 − 2q cos α + q 2

Then
cos α − q q sin α
U =q· and V = .
1 − 2q cos α + q 2 1 − 2q cos α + q 2

Exercise 8.8 Study the convergence of the series

∞
X
sin nx.
n=1

Solution to Exercise 8.8. Suppose that x ̸= kπ, where k ∈ Z and the series converges. Then due to
the n-th term test we get that
lim sin nx = 0, x ̸= kπ. (8.3)
n→∞

Then we have that

lim sin(n + 1)x = 0
n→∞
or  
lim sin nx cos x + cos nx sin x = 0.
n→∞

22
Taking into account (8.3), from the last relation we find that

lim cos nx = 0, x ̸= kπ. (8.4)

n→∞

Now from (8.3), (8.4) we obtain that

 
2 2
lim sin nx + cos nx = 0.
n→∞

This relation contradicts to the well known relation sin2 α + cos2 α = 1 and if x ̸= kπ, the series
diverges. If x = kπ the series converges and its sum is zero.

∞
X ∞
X
Exercise 8.9 Consider the series an with an ⩾ 0. Then the series a2n is also convergent.
n=1 n=1

∞
X
Solution to Exercise 8.9. It is evident that the sequence of partial sums of the series a2n is not
n=1
∞
X
decreasing. Moreover, due to the condition an ⩾ 0 and the convergence of series an , we have that
n=1

 2
e 2 2 2
An = a1 + a2 + ... + an < a1 + a2 + ... + an = A2n ⩽ const,

∞
X ∞
X
where An is a partial sum of an and A
en is a partial sum of a2n . Then by the monotone
n=1 n=1
convergence theorem there exists lim A
en and this series converges.
n→∞

Notice that the converse statement is not true. For example, if

1
an = ,
2n − 1
∞ ∞
X 1 X 1
the series 2
converges. However, diverges.
n=1
(2n − 1) n=1
(2n − 1)

Exercise 8.10 Suppose the series

∞
X ∞
X
def def
Sa = a2n and Sb = b2n
n=1 n=1

converge. Show that the series

∞ ∞ ∞
def
X def
X
2 def
X |an |
S1 = an b n , S2 = an + b n , S3 =
n=1 n=1 n=1
n

also converge.

23
Solution to Exercise 8.10. Consider the series S1 . We make use of the following inequality
1 2
an + b2n


an b n ⩽
2
and the convergence of series Sa , Sb , we have :
∞ ∞ ∞
!
X 1 X X 1
an b n = a2n + b2n = (Sa + Sb ) .
n=1
2 n=1 n=1
2

This means that the series S1 converges.

Consider the series S2 . Then it is easy to see that
∞ ∞ ∞ ∞ ∞
!
X
2 X X X X
an + b n = a2n + 2 an b n + b2n ⩽ 2 Sa + Sb + an b n ⩽
n=1 n=1 n=1 n=1 n=1
 
1
⩽ 2 Sa + Sb + (Sa + Sb ) .
2
This inequality implies that the series S2 also converges.
Consider the series S3 . We set bn = 1/n. Then taking into account the convergence of
∞
X 1
n=1
n2

and series S1 we obtain that the series S3 converges.

Exercise 8.11 Study the convergence of the series

∞
X 3n
.
n=1
n2

Solution to Exercise 8.11. We make use of d’Alembert’s test. We have :

3n+1 "  2 #
(n + 1)2
 n+1
n2

an+1 3 n
lim = lim = lim · = lim 3 =
n→∞ an n→∞ 3n n→∞ 3n (n + 1)2 n→∞ n+1
n2
 2  2
n+1−1 1
= 3 lim = 3 lim 1 − = 3.
n→∞ n+1 n→∞ n+1
Then this series diverges.

Exercise 8.12 Study the convergence of series

∞
X n3
.
n=1
(ln 3)n

24
Solution to Exercise 8.12. We make use of d’Alembert’s test. We have :

(n + 1)3 " #
an+1 (ln 3)n+1 (ln 3)n (n + 1)3
lim = lim = lim · =
n→∞ an n→∞ n3 n→∞ (ln 3)n+1 n3
(ln 3)n
"  3 #  3
1 n+1 1 1 1 1
= lim · = lim 1 + = ·1= .
n→∞ ln 3 n ln 3 n→∞ n ln 3 ln 3
Since ln 3 > ln e = 1 and (1/ ln 3) < 1, then the series converges.

Exercise 8.13 Study the convergence of the series

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

+ + ... + + ...
2! 4! (2n)!

Solution to Exercise 8.13. Using the d’Alembert test we calculate the following limit :

((n + 1)!)2 " #

an+1 (2 (n + 1))! ((n + 1)!)2 (2n)!
lim = lim = lim · =
n→∞ an n→∞ (n!)2 n→∞ (n!)2 (2n + 2)!
(2n)!

" # " #
(n! (n + 1))2 (2n)! (n!)2 (n + 1)2 1
= lim · = lim · =
n→∞ (n!)2 (2n)! (2n + 1) (2n + 2) n→∞ (n!)2 (2n + 1) (2n + 2)

(n + 1)2 n2 + 2n + 1 n2 + 2n + 1
= lim = lim = lim .
n→∞ (2n + 1) (2n + 2) n→∞ 4n2 + 2n + 4n + 2 n→∞ 4n2 + 6n + 2

Then
n2 + 2n + 1
2 1
an+1 n2 1+ n
+ n2 1
lim = lim 2 = lim 6 2 = < 1.
n→∞ an n→∞ 4n + 6n + 2 n→∞ 4 + + 4
n n2
n 2

This implies that the series converges.

Exercise 8.14 Determine whether the series

∞
X nn
3n−1
.
n=1
2

converges or diverges.

25
Solution to Exercise 8.14. We make use of the Cauchy test. We have
1
√
r
n nn n n n · 2n n n
lim 3n−1
= lim 3− 1 = lim 1 = lim = lim · lim 2 = ∞ · 1 = ∞ > 1.
n→∞ 2 n→∞ 2 n n→∞ 23 · 2− n n→∞ 8 n→∞ 8 n→∞

This implies that the series diverges.

Exercise 8.15 Determine whether the series

∞  n
X 1 1
+ .
n=1
3 n

converges or diverges.

Solution to Exercise 8.15. We make use of the Cauchy test. We have

s n  
n 1 1 1 1 1 1
lim + = lim + = + 0 = < 1.
n→∞ 3 n n→∞ 3 n 3 3

This implies that the series converges.

Exercise 8.16 Determine whether the series

1 1 1
2 + 3 + + ...
(ln 2) (ln 3) (ln 4)4

converges or diverges.

Solution to Exercise 8.16. The n-th term of series has the form :

1
an = .
(ln n)n
Then we make use of the Cauchy test to have
s
√ 1 1
lim n an = lim n n = lim = 0 < 1.
n→∞ n→∞ (ln n) n→∞ ln n

This implies that the series converges.

Exercise 8.17 Determine whether the series

∞
X sin2 n
(−1)n .
n=1
n

converges or diverges.

26
Solution to Exercise 8.17. Let us make use of Leibniz’s Theorem. Since sin2 n ⩽ 1, we have :

sin2 n sin2 n
lim |an | = lim = lim = 0.
n→∞ n→∞ n n→∞ n

This implies that the series converges.

Exercise 8.18 Determine whether the series

∞
X 2n + 1
(−1)n .
n=1
3n + 2

converges or diverges.

Solution to Exercise 8.18. Let us try to apply Leibniz’s Theorem. Then we will have :
2n+1 1
2n + 1 n
2+ n 2
lim |an | = lim = lim 3n+2 = lim 2 = ̸= 0.
n→∞ n→∞ 3n + 2 n→∞ n→∞ 3 + 3
n n

Then the absolute value of the n-th term does not go to zero as n → ∞. The series diverges.

Exercise 8.19 Study the convergence properties (absolute, conditional convergence) of the following
series :
∞
X (−1)n+1
.
n=1
n!

Solution to Exercise 8.19. We apply the d’Alembert test to the series determined by the absolute
values of the corresponding terms of our initial series. We have:
1
an+1 (n + 1)! n! 1
lim = lim = lim = lim = 0.
n→∞ an n→∞ 1 n→∞ (n + 1)! n→∞ n + 1
n!
This means that this series converges absolutely.

Exercise 8.20 Study the convergence properties (absolute, conditional convergence) of the following
series :
∞ √
X (−1)n+1 n
n=2
ln n

27
Solution to Exercise 8.20. First, let us find
√
n
lim |an | = lim .
n→∞ n→∞ ln n

We calculate this limit using L’Hospital’s rule :

1
√ √ √
n x 2 x 1 x 1 √
lim ∼ lim = lim = lim √ = lim x = ∞.
n→∞ ln n x→∞ ln x x→∞ 1 2 x→∞ x 2 x→∞
x
Thus the series is divergent.

Exercise 8.21 Study the convergence of the following series :

2 22 23 24
− + − + ...
3! 5! 7! 9!
Solution to Exercise 8.21. The n-th term of this series is given by

2n
an = (−1)n+1 .
(2n + 1)!
Let us apply d’Alembert test to the series
∞ ∞
X X 2n
|an | = .
n=1 n=1
(2n + 1)!

We have :
2n+1
an+1 (2n + 3)! 2n+1 (2n + 1)! 2
lim = lim n = lim = lim = 0.
n→∞ an n→∞ 2 n
n→∞ 2 (2n + 3)! n→∞ (2n + 2) (2n + 3)
(2n + 1)!
Then our series is absolutely convergent.

Exercise 8.22 Study the convergence properties (absolute, conditional convergence) of the following
series :
∞
X (−1)n+1
.
n=1
5n − 1

Solution to Exercise 8.22. We make use of Leibniz’s Theorem. First, we have :

1 1
lim |an | = lim = lim = 0.
n→∞ n→∞ 5n − 1 n→∞ 5n − 1

28
Thus the series converges.
Consider now the convergence of the series
∞ ∞
X (−1)n+1 X 1
= ,
n=1
5n − 1 n=1
5n − 1

In order to show the divergence of the last series we make use of the integral test. We have :
Z∞ Zn  n
dx dx 1 1
= lim = lim (ln |5x − 1|) = lim [ln (5n − 1) − ln 4] = ∞.
5x − 1 n→∞ 5x − 1 n→∞ 5 1 5 n→∞
1 1

This means that the our initial series converges only conditionally.

Exercise 8.23 Study the convergence properties (absolute, conditional convergence) of the following
series :
∞
X (−1)n
p .
n=1 n (n + 1)

Solution to Exercise 8.23. First, we apply Leibniz’s Theorem. We have :

1
lim |an | = lim p = 0.
n→∞ n→∞ n (n + 1)

Now we make use of the comparison test for the series :

∞ ∞ ∞
X X 1 X 1
|an | = p and .
n=1 n=1 n (n + 1) n=1
n

We have
√ 1 √ r s
n(n+1) n n2 n2 1
lim 1 = lim p = lim p = lim = lim 1 = 1.
n→∞
n
n→∞ n (n + 1) n→∞ n (n + 1) n→∞ n2 + n n→∞ 1+ n

Thus the series ∞ ∞

X X 1
|an | = p
n=1 n=1 n (n + 1)
is divergent and the initial series converges only conditionally.

Exercise 8.24 Show the convergence of the alternating series :

∞
X ln2 n
(−1)n−1 .
n=1
n

29
Solution to Exercise 8.24. We will apply Leibniz’s Theorem. To this end let us consider the function

ln2 x
def
φ(x) = .
x
We observe that by l’Hospital’s rule
lim φ(x) = 0
x→+∞

and
ln x
φ′ (x) = (2 − ln x).
x2
ln2 n
This implies that φ′ (x) < 0 if x > e2 . Thus we can conclude that the sequence an = satisfies
n
the conditions of Leibniz’s Theorem (n > e2 ) and the series converges.

Exercise 8.25 Show that if the sequence {an } goes to zero monotonically, then the series
∞
X
an sin nα
n=1

converges for all α ∈ R.

Solution to Exercise 8.25. Let us introduce the notation :

n
X
def
Bn = sin kα.
k=1

In order to find Bn we make use of the formula :

   
α 1 1
2 sin kα · sin = cos k − α − cos k + α.
2 2 2
Then one can show that
α nα
sin(n + 1) · sin
Bn = 2 2
α
sin
2
Then if α ̸= 2πm (m ∈ Z),
1
|Bn | ⩽ α
sin
2
and by Dirichlet test the series is convergent.

Exercise 8.26 Show that if the sequence {an } goes to zero monotonically, then the series
∞
X
an cos nα
n=1

converges for all α ̸= 2πm (m ∈ Z).

30
Solution to Exercise 8.25. In this exercise we proceed as in the previous one. Denote
n
X
def
Cn = cos kα.
k=1

Then using the relation

   
α 1 1
2 cos kα · sin = sin k + α − sin k − α,
2 2 2

we obtain that
1
Cn ⩽ α
sin
2
and by Dirichlet test the series is convergent.
If α = 2πm (m ∈ Z), we have that
∞
X ∞
X
an cos nα = an
n=1 n=1

and this series can be convergent or divergent.

31