CHAPTER 4

Series

Given a sequence a n , in many contexts it is natural to ask about the sum of

all the numbers in the sequence. If only a finite number of the a n are nonzero,
this is trivialand not very interesting. If an infinite number of the terms arent
zero, the path becomes less obvious. Indeed, its even somewhat questionable
whether it makes sense at all to add an infinite number of numbers.
There are many approaches to this question. The method given below is the
most common technique. Others are mentioned in the exercises.

1. What is a Series?
The idea behind adding up an infinite collection of numbers is a reduction to
the well-understood idea of a sequence. This is a typical approach in mathemat-
ics: reduce a question to a previously solved problem.

D EFINITION 4.1. Given a sequence a n , the series having a n as its terms is the
new sequence
n
X
sn = ak = a1 + a2 + + an .
k=1

The numbers s n are called the partial sums of the series. If s n ! S 2 R, then the
series converges to S. This is normally written as
1
X
a k = S.
k=1

Otherwise, the series diverges.

P
The notation 1 n=1 a n is understood to stand for the sequence of partial sums
of the series with terms a n . When there is no ambiguity, this is often abbreviated
P
to just a n .

E XAMPLE 4.1. If a n = (1)n for n 2 N, then s 1 = 1, s 2 = 1 + 1 = 0, s 3 =

1 + 1 1 = 1 and in general
(1)n 1
sn =
2
does not converge because it oscillates between 1 and 0. Therefore, the series
P
(1)n diverges.

4-1
4-2 Series

E XAMPLE 4.2 (Geometric Series). Recall that a sequence of the form a n =

n1
cr is called a geometric sequence. It gives rise to a series
1
X
c r n1 = c + cr + cr 2 + cr 3 +
n=1

called a geometric series. The number r is called the ratio of the series.
Suppose a n = r n1 for r 6= 1. Then,
s 1 = 1, s 2 = 1 + r, s 3 = 1 + r + r 2 , . . .
In general, it can be shown by induction (or even long division of polynomials)
that
Xn Xn 1rn
(4.1) sn = ak = r k1 = .
k=1 k=1 1r
The convergence of s n in (4.1) depends on the value of r . Letting n ! 1, its
apparent that s n diverges when |r | > 1 and converges to 1/(1 r ) when |r | < 1.
When r = 1, s n = n ! 1. When r = 1, its essentially the same as Example 4.1,
and therefore diverges. In summary,
1
X c
c r n1 =
n=1 1r
for |r | < 1, and diverges when |r | 1. This is called a geometric series with ratio r .

F IGURE 4.1. Stepping to the wall.

steps

2 1 1/2 0
distance from wall

In some cases, the geometric series has an intuitively plausible limit. If you
start two meters away from a wall and keep stepping halfway to the wall, no
number of steps will get you to the wall, but a large number of steps will get you
as close to the wall as you want. (See Figure 4.1.) So, the total distance stepped
has limiting value 2. The total distance after n steps is the nth partial sum of a
geometric series with ratio r = 1/2 and c = 1.
P
E XAMPLE 4.3 (Harmonic Series). The series 1 n=1 1/n is called the harmonic
series. It was shown in Example 3.18 that the harmonic series diverges.

E XAMPLE 4.4. The terms of the sequence

1
an = , n 2 N.
n2 + n
June 23, 2017 http://math.louisville.edu/lee/ira
Basic Definitions 4-3

can be decomposed into partial fractions as

1 1
an = .
n n +1
If s n is the series having a n as its terms, then s 1 = 1/2 = 1 1/2. We claim that
s n = 1 1/(n + 1) for all n 2 N. To see this, suppose s k = 1 1/(k + 1) for some
k 2 N. Then

1 1 1 1
s k+1 = s k + a k+1 = 1 + = 1
k +1 k +1 k +2 k +2
and the claim is established by induction. Now its easy to see that
1
X 1 1
2
= lim 1 = 1.
n=1 n + n n +2
n!1

This is an example of a telescoping series. The name is apparently based on the

idea that the middle terms of the series cancel, causing the series to collapse like
a hand-held telescope.

The following theorem is an easy consequence of the properties of sequences

shown in Theorem 3.8.
P P
T HEOREM 4.2. Let a n and b n be convergent series.
P P
(a) If c 2 R, then c a n = c a n .
P P P
(b) (a n + b n ) = a n + b n .
(c) a n ! 0
P P
P ROOF. Let A n = nk=1 a k and B n = nk=1 b k be the sequences of partial sums
for each of the two series. By assumption, there are numbers A and B where
A n ! A and B n ! B .
P P
(a) nk=1 c a k = c nk=1 a k = c A n ! c A.
P P P
(b) nk=1 (a k + b k ) = nk=1 a k + nk=1 b k = A n + B n ! A + B .
P P
(c) For n > 1, a n = nk=1 a k n1 a = A n A n1 ! A A = 0.
k=1 k

Notice that the first two parts of Theorem 4.2 show that the set of all conver-
gent series is closed under linear combinations.
Theorem 4.2(c) is very useful because its contrapositive provides the most
basic test for divergence.
P
C OROLLARY 4.3 (Going to Zero Test). If a n 6! 0, then a n diverges.

Many have made the mistake of reading too much into Corollary 4.3. It can
only be used to show divergence. When the terms of a series do tend to zero, that
does not guarantee convergence. Example 4.3, shows Theorem 4.2(c) is necessary,
but not sufficient for convergence.
Another useful observation is that the partial sums of a convergent series are
a Cauchy sequence. The Cauchy criterion for sequences can be rephrased for
series as the following theorem, the proof of which is Exercise 4.4.

June 23, 2017 http://math.louisville.edu/lee/ira

4-4 Series

P
T HEOREM 4.4 (Cauchy Criterion for Series). Let a n be a series. The following
statements are equivalent.
P
(a) a n converges.
(b) For every " > 0 there is an N 2 N such that whenever n m N , then

X n

a i < ".
i =m

2. Positive Series
Most of the time, it is very hard or impossible to determine the exact limit of
a convergent series. We must satisfy ourselves with determining whether a series
converges, and then approximating its sum. For this reason, the study of series
usually involves learning a collection of theorems that might answer whether a
given series converges, but dont tell us to what it converges. These theorems are
usually called the convergence tests. The reader probably remembers a battery
of such tests from her calculus course. There is a myriad of such tests, and the
standard ones are presented in the next few sections, along with a few of those
less widely used.
Since convergence of a series is determined by convergence of the sequence
of its partial sums, the easiest series to study are those with well-behaved partial
sums. Series with monotone sequences of partial sums are certainly the simplest
such series.
P
D EFINITION 4.5. The series a n is a positive series, if a n 0 for all n.

The advantage of a positive series is that its sequence of partial sums is

nonnegative and increasing. Since an increasing sequence converges if and only
if it is bounded above, there is a simple criterion to determine whether a positive
series converges. All of the standard convergence tests for positive series exploit
this criterion.

2.1. The Most Common Convergence Tests. All beginning calculus courses
contain several simple tests to determine whether positive series converge. Most
of them are presented below.
2.1.1. Comparison Tests. The most basic convergence tests are the compari-
son tests. In these tests, the behavior of one series is inferred from that of another
series. Although theyre easy to use, there is one often fatal catch: in order to use
a comparison test, you must have a known series to which you can compare the
mystery series. For this reason, a wise mathematician collects example series
for her toolbox. The more samples in the toolbox, the more powerful are the
comparison tests.
P P
T HEOREM 4.6 (Comparison Test). Suppose a n and b n are positive series
with a n b n for all n.
P P
(a) If b n converges, then so does a n .
P P
(b) If a n diverges, then so does b n .

June 23, 2017 http://math.louisville.edu/lee/ira

Positive Series 4-5

a 1 + a 2 + a 3 + a 4 + a 5 + a 6 + a 7 + a 8 + a 9 + + a 15 +a 16 +
| {z } | {z } | {z }
2a 2 4a 4 8a 8

a 1 + a 2 + a 3 + a 4 + a 5 + a 6 + a 7 + a 8 + a 9 + + a 15 +a 16 +
|{z} | {z } | {z } | {z }
a 2 2a 4 4a 8 8a 16

F IGURE 4.2. This diagram shows the groupings used in inequality (4.3).

P P
P ROOF. Let A n and B n be the partial sums of a n and b n , respectively. It
follows from the assumptions that A n and B n are increasing and for all n 2 N,
(4.2) An Bn .
P P
If b n = B , then (4.2) implies B is an upper bound for A n , and a n con-
verges.
P
On the other hand, if a n diverges, A n ! 1 and the Sandwich Theorem
3.9(b) shows B n ! 1.
P
E XAMPLE 4.5. Example 4.3 shows that 1/n diverges. If p 1, then 1/n p
P
1/n, and Theorem 4.6 implies 1/n p diverges.
P
E XAMPLE 4.6. The series sin2 n/2n converges because
sin2 n 1
n
2n 2
P
for all n and the geometric series 1/2n = 1.

T HEOREM 4.7 (Cauchys Condensation Test1). Suppose a n is a decreasing

sequence of nonnegative numbers. Then
X X
a n converges iff 2n a 2n converges.

P ROOF. Since a n is decreasing, for n 2 N,

2n+1
X1 n
2X 1
(4.3) a k 2n a 2n 2 ak .
k=2n k=2n1

(See Figure 2.1.1.) Adding for 1 n m gives

2m+1
X1 m
X 2m
X 1
k
(4.4) ak 2 a 2k 2 ak .
k=2 k=1 k=1
P P
Suppose a n converges to S. The right-hand inequality of (4.4) shows m
k=1
2 k a 2k <
P k P
2S and 2 a 2k must converge. On the other hand, if a n diverges, then the left-
P
hand side of (4.4) is unbounded, forcing 2k a 2k to diverge.

1The series P 2n a P
2n is sometimes called the condensed series associated with an .

June 23, 2017 http://math.louisville.edu/lee/ira

4-6 Series

P
E XAMPLE 4.7 (p-series). For fixed p 2 R, the series 1/n p is called a p-series.
The special case when p = 1 is the harmonic series. Notice
X 2n X 1p n
= 2
(2n )p
is a geometric series with ratio 21p , so it converges only when 21p < 1. Since
21p < 1 only when p > 1, it follows from the Cauchy Condensation Test that
the p-series converges when p > 1 and diverges when p 1. (Of course, the
divergence half of this was already known from Example 4.5.)
The p-series are often useful for the Comparison Test, and also occur in many
areas of advanced mathematics such as harmonic analysis and number theory.
P P
T HEOREM 4.8 (Limit Comparison Test). Suppose a n and b n are positive
series with
an an
(4.5) = lim inf lim sup = .
bn bn
P P P
(a) If 2 (0, 1) and a n converges, then so does b n , and if b n di-
P
verges, then so does a n .
P P P
(b) If 2 (0, 1) and b n diverges, then so does a n , and if a n con-
P
verges, then so does b n .

P ROOF. To prove (a), suppose > 0. There is an N 2 N such that

an
(4.6) n N =) < .
2 bn
If n > N , then (4.6) gives
Xn n
X
(4.7) bk < ak
2 k=N k=N
P P
If a n converges, then (4.7) shows the partial sums of b n are bounded and
P P P
b n converges. If b n diverges, then (4.7) shows the partial sums of a n are
P
unbounded, and a n must diverge.
The proof of (b) is similar.
The following easy corollary is the form this test takes in most calculus books.
Its easier to use than Theorem 4.8 and suffices most of the time.
P P
C OROLLARY 4.9 (Limit Comparison Test). Suppose a n and b n are positive
series with
an
(4.8) = lim .
n!1 b n
P P
If 2 (0, 1), then a n and b n either both converge or both diverge.
X 1
E XAMPLE 4.8. To test the series for convergence, let
2n n
1 1
an = and b n = .
2n n 2n
June 23, 2017 http://math.louisville.edu/lee/ira
Positive Series 4-7

Then
an 1/(2n n) 2n 1
lim= lim n
= lim n
= lim = 1 2 (0, 1).
n!1 b n n!1 1/2 n!1 2 n n!1 1 n/2n
P
Since 1/2n = 1, the original series converges by the Limit Comparison Test.

2.1.2. Geometric Series-Type Tests. The most important series is undoubt-

edly the geometric series. Several standard tests are basically comparisons to
geometric series.
P
T HEOREM 4.10 (Root Test). Suppose a n is a positive series and

= lim sup a n1/n .

P P
If < 1, then a n converges. If > 1, then a n diverges.

P ROOF. First, suppose < 1 and r 2 (, 1). There is an N 2 N so that a n1/n < r
for all n N . This is the same as a n < r n for all n N . Using this, it follows that
when n N ,
n
X NX
1 n
X NX
1 n
X NX
1 rN
ak = ak + ak < ak + rk < ak + .
k=1 k=1 k=N k=1 k=N k=1 1r
P
This shows the partial sums of a n are bounded. Therefore, it must converge.
1/k
If > 1, there is an increasing sequence of integers k n ! 1 such that a k n > 1
P n
for all n 2 N. This shows a kn > 1 for all n 2 N. By Theorem 4.3, a n diverges.
P
E XAMPLE 4.9. For any x 2 R, the series |x n |/n! converges. To see this, note
that according to Exercise 3.3.7,
1/n
|x n | |x|
= ! 0 < 1.
n! (n!)1/n
Applying the Root Test shows the series converges.
P P
E XAMPLE 4.10. Consider the p-series 1/n and 1/n 2 . The first diverges
and the second converges. Since n 1/n ! 1 and n 2/n ! 1, it can be seen that when
= 1, the Root Test in inconclusive.
P
T HEOREM 4.11 (Ratio Test). Suppose a n is a positive series. Let
a n+1 a n+1
r = lim inf
lim sup = R.
an an
P P
If R < 1, then a n converges. If r > 1, then a n diverges.

P ROOF. First, suppose R < 1 and 2 (R, 1). There exists N 2 N such that
a n+1 /a n < whenever n N . This implies a n+1 < a n whenever n N . From
this its easy to prove by induction that a N +m < m a N whenever m 2 N. It follows

June 23, 2017 http://math.louisville.edu/lee/ira

4-8 Series

that, for n > N ,

n
X N
X n
X
ak = ak + ak
k=1 k=1 k=N +1
N
X nN
X
= ak + a N +k
k=1 k=1
N
X nN
X
< ak + aN k
k=1 k=1
N
X aN
< ak + .
k=1 1
P P
Therefore, the partial sums of a n are bounded, and a n converges.
If r > 1, then choose N 2 N so that a n+1 > a n for all n N . Its now apparent
that a n 6! 0.

In calculus books, the ratio test usually takes the following simpler form.
P
C OROLLARY 4.12 (Ratio Test). Suppose a n is a positive series. Let
a n+1
r = lim
.
n!1 a n
P P
If r < 1, then a n converges. If r > 1, then a n diverges.

From a practical viewpoint, the ratio test is often easier to apply than the root
test. But, the root test is actually the stronger of the two in the sense that there
are series for which the ratio test fails, but the root test succeeds. (See Exercise
4.10, for example.) This happens because
a n+1 a n+1
(4.9) lim inf lim inf a n1/n lim sup a n1/n lim sup .
an an
To see this, note the middle inequality is always true. To prove the right-hand
inequality, choose r > lim sup a n+1 /a n . It suffices to show lim sup a n1/n r . As in
the proof of the ratio test, a n+k < r k a n . This implies
an
a n+k < r n+k ,
rn
which leads to
a 1/(n+k)
1/(n+k) n
a n+k <r n
.
r
Finally,
a 1/(n+k)
1/(n+k) n
lim sup a n1/n = lim sup a n+k lim sup r = r.
k!1 k!1 rn
The left-hand inequality is proved similarly.

June 23, 2017 http://math.louisville.edu/lee/ira

Kummer-Type Tests 4-9

2.2. Kummer-Type Tests.

This is an advanced section that can be omitted.

Most times the simple tests of the preceding section suffice. However, more
difficult series require more delicate tests. There dozens of other, more spe-
cialized, convergence tests. Several of them are consequences of the following
theorem.
P
T HEOREM 4.13 (Kummers Test). Suppose a n is a positive series, p n is a
sequence of positive numbers and

an an
(4.10) = lim inf p n p n+1 lim sup p n p n+1 =
a n+1 a n+1
P P P
If > 0, then a n converges. If 1/p n diverges and < 0, then a n diverges.
P
P ROOF. Let s n = nk=1 a k , suppose > 0 and choose r 2 (0, ). There must be
an N > 1 such that
an
pn p n+1 > r, 8n N .
a n+1
Rearranging this gives

(4.11) p n a n p n+1 a n+1 > r a n+1 , 8n N .

For M > N , (4.11) implies

M
X XM
p n a n p n+1 a n+1 > r a n+1
n=N n=N
p N a N p M +1 a M +1 > r (s M s N 1 )
p N a N p M +1 a M +1 + r s N 1 > r s M
p N a N + r s N 1
> sM
r
P
Since N is fixed, the left side is an upper bound for s M , and it follows that a n
converges.
P
Next suppose 1/p n diverges and < 0. There must be an N 2 N such that
an
pn p n+1 < 0, 8n N .
a n+1
This implies
p n a n < p n+1 a n+1 , 8n N .
Therefore, p n a n > p N a N whenever n > N and
1
an > p N a N , 8n N .
pn
P P
Because N is fixed and 1/p n diverges, the Comparison Test shows a n diverges.

June 23, 2017 http://math.louisville.edu/lee/ira

4-10 Series

Kummers test is powerful. In fact, it can be shown that, given any positive
series, a judicious choice of the sequence p n can always be made to determine
whether it converges. (See Exercise 4.17, [20] and [19].) But, as stated, Kummers
test is not very useful because choosing p n for a given series is often difficult.
Experience has led to some standard choices that work with large classes of series.
For example, Exercise 4.9 asks you to prove the choice p n = 1 for all n reduces
Kummers test to the standard ratio test. Other useful choices are shown in the
following theorems.
P
T HEOREM 4.14 (Raabes Test). Let a n be a positive series such that a n > 0 for
all n. Define

an an
= lim sup n 1 lim inf n 1 =
n!1 a n+1 n!1 a n+1
P P
If > 1, then a n converges. If < 1, then a n diverges.

P ROOF. Let p n = n in Kummers test, Theorem 4.13.

When Raabes test is inconclusive, there are even more delicate tests, such as
the theorem given below.
P
T HEOREM 4.15 (Bertrands Test). Let a n be a positive series such that a n > 0
for all n. Define

an an
= lim inf ln n n 1 1 lim sup ln n n 1 1 = .
n!1 a n+1 n!1 a n+1
P P
If > 1, then a n converges. If < 1, then a n diverges.

P ROOF. Let p n = n ln n in Kummers test.

E XAMPLE 4.11. Consider the series
!p
X X Y n 2k
(4.12) an = .
k=1 2k + 1

Its of interest to know for what values of p it converges.

An easy computation shows that a n+1 /a n ! 1, so the ratio test is inconclusive.
Next, try Raabes test. Manipulating
2n+3 p
an p 1
lim n 1 = lim 2n+21
n!1 a n+1 n!1
n

it becomes a 0/0 form and can be evaluated with LHospitals rule.2

n p
n 2 3+2
2+2 n p p
lim = .
n!1 (1 + n) (3 + 2 n) 2
From Raabes test, Theorem 4.14, it follows that the series converges when p > 2
and diverges when p < 2. Raabes test is inconclusive when p = 2.

2See 5.2.

June 23, 2017 http://math.louisville.edu/lee/ira

Absolute and Conditional Convergence 4-11

Now, suppose p = 2. Consider

an (4 + 3 n)
lim ln n n 1 1 = lim ln n =0
n!1 a n+1 n!1 4 (1 + n)2
and Bertrands test, Theorem 4.15, shows divergence.
The series (4.12) converges only when p > 2.

3. Absolute and Conditional Convergence

The tests given above are for the restricted case when a series has positive
terms. If the stipulation that the series be positive is thrown out, things becomes
considerably more complicated. But, as is often the case in mathematics, some
problems can be attacked by reducing them to previously solved cases. The
following definition and theorem show how to do this for some special cases.
P P P
D EFINITION 4.16. Let a n be a series. If |a n | converges, then a n is
absolutely convergent. If it is convergent, but not absolutely convergent, then it is
conditionally convergent.
P
Since |a n | is a positive series, the preceding tests can be used to deter-
mine its convergence. The following theorem shows that this is also enough for
convergence of the original series.
P
T HEOREM 4.17. If a n is absolutely convergent, then it is convergent.
P ROOF. Let " > 0. Theorem 4.4 yields an N 2 N such that when n m N ,

n
X X n

"> |a k | a k 0.
k=m
k=m

Another application Theorem 4.4 finishes the proof.
P
E XAMPLE 4.12. The series (1)n+1 /n is called the alternating harmonic
series. (See Figure 4.3.) Since the harmonic series diverges, we see the alternating
harmonic series is not absolutely convergent.
P
On the other hand, if s n = nk=1 (1)k+1 /k, then
Xn 1 1

Xn 1
s 2n = =
k=1 2k 1 2k k=1 2k(2k 1)
is a positive series that converges by the Comparison Test. Since |s 2n s 2n1 | =
1/2n ! 0, its clear that s 2n1 must also converge to the same limit. Therefore, s n
P
converges and (1)n+1 /n is conditionally convergent. (Another way to show
the alternating harmonic series converges is shown in Example 3.18.)
To summarize: absolute convergence implies convergence, but convergence
does not imply absolute convergence.
There are a few tests that address conditional convergence. Following are the
most well-known.
T HEOREM 4.18 (Abels Test). Let a n and b n be sequences satisfying

June 23, 2017 http://math.louisville.edu/lee/ira

4-12 Series

1.0

0.8

0.6

0.4

0.2

0 5 10 15 20 25 30 35

F IGURE 4.3. This plot shows the first 35 partial sums of the alternating
harmonic series. It can be shown it converges to ln 2 0.6931, which is
the level of the dashed line. Notice how the odd partial sums decrease
to ln 2 and the even partial sums increase to ln 2.

P
(a) s n = nk=1 a k is a bounded sequence.
(b) b n b n+1 , 8n 2 N
(c) b n ! 0
P
Then a n b n converges.

To prove this theorem, the following lemma is needed.

L EMMA 4.19 (Summation by Parts). For every pair of sequences a n and b n ,

n
X n
X n
X k
X
a k b k = b n+1 ak (b k+1 b k ) a`
k=1 k=1 k=1 `=1
Pn
P ROOF. Let s 0 = 0 and s n = k=1
a k when n 2 N. Then

n
X n
X
ak bk = (s k s k1 )b k
k=1 k=1
Xn n
X
= sk bk s k1 b k
k=1 k=1
!
n
X n
X
= sk bk s k b k+1 s n b n+1
k=1 k=1
n
X n
X k
X
= b n+1 ak (b k+1 b k ) a`
k=1 k=1 `=1

June 23, 2017 http://math.louisville.edu/lee/ira

Absolute and Conditional Convergence 4-13
P
P ROOF. To prove the theorem, suppose nk=1 a k < M for all n 2 N. Let " > 0
and choose N 2 N such that b N < "/2M . If N m < n, use Lemma 4.19 to write

Xn X n m1
X

a` b` = a` b` a` b`
`=m `=1 `=1


Xn Xn X

= b n+1 a` (b `+1 b ` ) ak
`=1 `=1 k=1
!
m1
X m1
X X

bm a` (b `+1 b ` ) ak
`=1 `=1 k=1

Using (a) gives

n
X
(b n+1 + b m )M + M |b `+1 b ` |
`=m

Now, use (b) to see

n
X
= (b n+1 + b m )M + M (b ` b `+1 )
`=m

and then telescope the sum to arrive at

= (b n+1 + b m )M + M (b m b n+1 )
= 2M b m
"
< 2M
2M
<"
Pn
This shows `=1
a ` b ` satisfies Theorem 4.4, and therefore converges.

Theres one special case of this theorem thats most often seen in calculus
texts.

C OROLLARY 4.20 (Alternating Series Test). If c n decreases to 0, then the series

P P
(1)n+1 c n converges. Moreover, if s n = nk=1 (1)k+1 c k and s n ! s, then |s n s| <
c n+1 .

P ROOF. Let a n = (1)n+1 and b n = c n in Theorem 4.18 to see the series con-
P
verges to some number s. For n 2 N, let s n = nk=0 (1)k+1 c k and s 0 = 0. Since

s 2n s 2n+2 = c 2n+1 + c 2n+2 0 and s 2n+1 s 2n+3 = c 2n+2 c 2n+3 0,

It must be that s 2n " s and s 2n+1 # s. For all n 2 !,

0 s 2n+1 s s 2n+1 s 2n+2 = c 2n+2 and 0 s s 2n s 2n+1 s 2n = c 2n+1 .

This shows |s n s| < c n+1 for all n.

June 23, 2017 http://math.louisville.edu/lee/ira

4-14 Series

F IGURE 4.4. Here is a more whimsical way to visualize the partial

sums of the alternating harmonic series.

A series such as that in Corollary 4.20 is called an alternating series. More

P
formally, if a n is a sequence such that a n /a n+1 < 0 for all n, then a n is an
alternating series. Informally, it just means the series alternates between positive
and negative terms.

E XAMPLE 4.13. Corollary 4.20 provides another way to prove the alternating
harmonic series in Example 4.12 converges. Figures 4.3 and 4.4 show how the
partial sums bounce up and down across the sum of the series.

4. Rearrangements of Series
This is an advanced section that can be omitted.

We want to use our standard intuition about adding lists of numbers when
working with series. But, this intuition has been formed by working with finite
sums and does not always work with series.
P P
E XAMPLE 4.14. Suppose (1)n+1 /n = so that (1)n+1 2/n = 2. Its easy
to show > 1/2. Consider the following calculation.

X 2
2 = (1)n+1
n
2 1 2 1
= 21+ + +
3 2 5 3

June 23, 2017 http://math.louisville.edu/lee/ira

Rearrangements of Series 4-15

Rearrange and regroup.

1 2 1 1 2 1 1
= (2 1) + + +
2 3 3 4 5 5 6
1 1 1
= 1 + +
2 3 4
=
So, = 2 with 6= 0. Obviously, rearranging and regrouping of this series is a
questionable thing to do.
In order to carefully consider the problem of rearranging a series, a precise
definition is needed.
P
D EFINITION 4.21. Let : N ! N be a bijection and a n be a series. The new
P
series a (n) is a rearrangement of the original series.
The problem with Example 4.14 is that the series is conditionally convergent.
Such examples cannot happen with absolutely convergent series. For the most
part, absolutely convergent series behave as we are intuitively led to expect.
P P
T HEOREM 4.22. If a n is absolutely convergent and a (n) is a rearrangement
P P P
of a n , then a (n) = a n .
P
P ROOF. Let a n = s and " > 0. Choose N 2 N so that N m < n implies
Pn
k=m k
|a | < ". Choose M N such that
{1, 2, . . . , N } {(1), (2), . . . , (M )}.
If P > M , then
XP XP 1
X

ak a (k) |a | "
k=1 k=1
k=N +1 k
and both series converge to the same number.
When a series is conditionally convergent, the result of a rearrangement is
hard to predict. This is shown by the following surprising theorem.
P
T HEOREM 4.23 (Riemann Rearrangement). If a n is conditionally convergent
P
and c 2 R [ {1, 1}, then there is a rearrangement such that a (n) = c.
To prove this, the following lemma is needed.
P
L EMMA 4.24. If a n is conditionally convergent and
( (
an , an > 0 a n , a n < 0
bn = and c n = ,
0, an 0 0, an 0
P P
then both b n and c n diverge.
P P
P ROOF. Suppose b n converges. By assumption, a n converges, so Theo-
rem 4.2 implies X X X
cn = bn an

June 23, 2017 http://math.louisville.edu/lee/ira

4-16 Series

converges. Another application of Theorem 4.2 shows

X X X
|a n | = b n + c n
P
converges. This is a contradiction of the assumption that a n is conditionally
P
convergent, so b n cannot converge.
P
A similar contradiction arises under the assumption that c n converges.

P ROOF. (Theorem 4.23) Let b n and c n be as in Lemma 4.24 and define the
subsequence a n+ of b n by removing those terms for which b n = 0 and a n 6= 0.
Define the subsequence a n of c n by removing those terms for which c n = 0. The
P P1
series 1 +
n=1 a n and n=1 a n are still divergent because only terms equal to zero
have been removed from b n and c n .
Now, let c 2 R and m 0 = n 0 = 0. According to Lemma 4.24, we can define the
natural numbers
Xn m1
X n
X
m 1 = min{n : a k+ > c} and n 1 = min{n : a k+ + a ` < c}.
k=1 k=1 `=1
If m p and n p have been chosen for some p 2 N, then define
( ! )
Xp mXk+1 nXk+1 Xn
+ +
m p+1 = min n : a` a` + a` > c
k=0 `=m k +1 `=n k +1 `=m p +1

and
( !
p
X mX
k+1 nX
k+1
n p+1 = min n : a `+ a `
k=0 `=m k +1 `=n k +1
nX
)
p+1 n
X
+ a `+ a ` <c .
`=m p +1 `=n p +1

Consider the series

(4.13) a 1+ + a 2+ + + a m
+
1
a 1 a 2 a n1
+ + +
+ am 1 +1
+ am 1 +2
+ + am 2
a n1 +1 a n1 +2 a n2
+ + +
+ am 2 +1
+ am 2 +2
+ + am 3
a n2 +1 a n2 +2 a n3
+
P
It is clear this series is a rearrangement of 1n=1 a n and the way in which m p and
n p were chosen guarantee that
m
!
p1
X mX k+1 nk
X Xp
+ + +
0< a` a` + ak c am p
k=0 `=m k +1 `=n k +1 k=m p +1

and !
p
X mX
k+1 nk
X
0<c a `+ a ` a np
k=0 `=m k +1 `=n k +1
+
Since both am
! 0 and
p
a n p ! 0, the result follows from the Squeeze Theorem.
The argument when c is infinite is left as Exercise 4.31.

June 23, 2017 http://math.louisville.edu/lee/ira

5. EXERCISES 4-17

A moral to take from all this is that absolutely convergent series are robust
and conditionally convergent series are fragile. Absolutely convergent series can
be sliced and diced and mixed with careless abandon without getting surprising
results. If conditionally convergent series are not handled with care, the results
can be quite unexpected.

5. Exercises

4.1. Prove Theorem 4.4.

P1 P1 1
4.2. If n=1 a n is a convergent positive series, then does n=1 1+a n converge?

1
X
4.3. The series (a n a n+1 ) converges iff the sequence a n converges.
n=1
P
4.4. Prove or give a counter example: If |a n | converges, then na n ! 0.

4.5. If the series a 1 + a 2 + a 3 + converges to S, then so does

(4.14) a1 + 0 + a2 + 0 + 0 + a3 + 0 + 0 + 0 + a4 + .

P
4.6. If 1 n=1 a n converges and b n is a bounded monotonic sequence, then
P1
n=1 a n b n converges.
P1 n
4.7. Let x n be a sequence with range {0, 1, 2, 3, 4, 5, 6, 7, 8, 9}. Prove that n=1 x n 10
converges and its sum is in the interval [0, 1].

4.8. Write 6.17272727272 as a fraction.

4.9. Prove the ratio test by setting p n = 1 for all n in Kummers test.

4.10. Consider the series

1 1 1 1
1+1+ + + + + = 4.
2 2 4 4
Show that the ratio test is inconclusive for this series, but the root test gives a
positive answer.
1
X 1
4.11. Does converge?
n=2 n(ln n)2

4.12. Does
1 12 123 1234
+ + + +
3 35 357 3579
converge?

June 23, 2017 http://math.louisville.edu/lee/ira

4-18 Series

4.13. For what values of p does

p
1 13 p 135 p
+ + +
2 24 246
converge?

4.14. Find sequences a n and b n satisfying:

(a) a n > 0, 8n 2 N and a n ! 0;
P
(b) B n = nk=1 b k is a bounded sequence; and,
P1
(c) n=1 a n b n diverges.

4.15. Let a n be a sequence such that a 2n ! A and a 2n a 2n1 ! 0. Then a n ! A.

4.16. Prove Bertrands test, Theorem 4.15.

P P
4.17. Let a n be a positive series. Prove that a n converges if and only if there
is a sequence of positive numbers p n and > 0 such that
an
lim p n p n+1 = .
n!1 a n+1
P Pn
(Hint: If s = a n and s n = k=1
a k , then let p n = (s s n )/a n .)
P1 xn
4.18. Prove that n=0 n! converges for all x 2 R.
P1
4.19. Find all values of x for which k=0
k 2 (x + 3)k converges.

4.20. For what values of x does the series

X1 (1)n+1 x 2n1
(4.19)
n=1 2n 1

converge?

X1 (x + 3)n
4.21. For what values of x does n
converge absolutely, converge
n=1 n4
conditionally or diverge?
1
X n +6
4.22. For what values of x does converge absolutely, converge
n=1 n 2 (x 1)n
conditionally or diverge?
P1 n
4.23. For what positive values of does n=1 n converge?
X n
4.24. Prove that cos sin converges.
3 n

June 23, 2017 http://math.louisville.edu/lee/ira

5. EXERCISES 4-19

P1
4.25. For a series k=1
a n with partial sums s n , define
1 Xn
n = sn .
n k=1
P
Prove that if 1 a = s, then n ! s. Find an example where n converges, but
k=1 n
P1
k=1
a n does not. (If n converges, the sequence is said to be Cesro summable.)
P
4.26. If a n is a sequence with a subsequence b n , then 1 n=1 b n is a subseries
P P1 P
of 1n=1 na . Prove that if every subseries of a
n=1 n converges, then 1 n=1 a n
converges absolutely.
P P1 2
4.27. If 1 n=1 a n is a convergent positive series, then so is n=1 a n . Give an
example to show the converse is not true.
P1 P1 2
4.28. Prove or give a counter example: If n=1 a n converges, then n=1 a n
converges.
P1 n
4.29. For what positive values of does n=1 n converge?

4.30. If a n 0 for all n 2 N and there is a p > 1 such that limn!1 n p a n exists and
P
is finite, then 1 n=1 a n converges. Is this true for p = 1?

4.31. Finish the proof of Theorem 4.23.

4.32. Leonhard Euler started with the equation

x x
+ = 0,
x 1 1x
transformed it to
1 x
+ = 0,
1 1/x 1 x
and then used geometric series to write it as
1 1
(4.23) + 2 + + 1 + x + x 2 + x 3 + = 0.
x x
Show how Euler did his calculation and find his mistake.
P
4.33. Let a n be a conditionally convergent series and c 2 R [ {1, 1}. There is
P
a sequence b n such that |b n | = 1 for all n 2 N and a n b n = c.

June 23, 2017 http://math.louisville.edu/lee/ira