10 INFINITE SERIES

10.1 Sequences (LT Section 11.1)

Preliminary Questions
1. What is a4 for the sequence an = n2 − n?
solution Substituting n = 4 in the expression for an gives

a4 = 42 − 4 = 12

2. Which of the following sequences converge to zero?

 n
n2 −1
(a) 2 (b) 2n (c)
n +1 2
solution
(a) This sequence does not converge to zero:
n2 x2 1 1
lim = lim = lim = =1
n→∞ n2 + 1 x→∞ x2 + 1 x→∞ 1 + 1
x2
1+0

(b) This sequence does not converge to zero: this is a geometric sequence with r = 2 > 1; hence, the
sequence diverges to ∞.
(c) Recall that if |an | converges to 0, then an must also converge to zero. Here,
 n   n
 1  1
 − =
2  2

which is a geometric sequence with 0 < r < 1; hence, ( 21 )n converges to zero. It therefore follows that (− 21 )n
converges to zero.
√
3. Let an be the nth decimal approximation to 2. That is, a1 = 1, a2 = 1.4, a3 = 1.41, and so on. What is
lim an ?
n→∞
√
solution lim an = 2
n→∞
4. Which of the following sequences is defined recursively?
√ 
(a) an = 4 + n (b) bn = 4 + bn−1
solution
(a) an can be computed directly, since it depends on n only and not on preceding terms. Therefore an is
defined explicitly and not recursively.
(b) bn is computed in terms of the preceding term bn−1 , hence the sequence {bn } is defined recursively.
5. Theorem 5 says that every convergent sequence is bounded. Determine if the following statements are
true or false, and if false, give a counterexample:
(a) If {an } is bounded, then it converges.
(b) If {an } is not bounded, then it diverges.
(c) If {an } diverges, then it is not bounded.
solution
(a) This statement is false. The sequence an = cos πn is bounded since −1 ≤ cos πn ≤ 1 for all n, but it does
not converge: since an = cos nπ = (−1)n , the terms assume the two values 1 and −1 alternately, hence they
do not approach one value.
(b) By Theorem 5, a converging sequence must be bounded. Therefore, if a sequence is not bounded, it
certainly does not converge.
(c) The statement is false. The sequence an = (−1)n is bounded, but it does not approach one limit.

1
2 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

Exercises
1. Match each sequence with its general term:

a1 , a2 , a3 , a4 , . . . General term

(a) 12 , 23 , 34 , 45 , . . . (i) cos πn

n!
(b) −1, 1, −1, 1, . . . (ii)
2n
(c) 1, −1, 1, −1, . . . (iii) (−1)n+1
n
(d) 12 , 24 , 68 , 24
16 . . . (iv)
n+1

solution
(a) The numerator of each term is the same as the index of the term, and the denominator is one more than
the numerator; hence an = n+1
n
, n = 1, 2, 3, . . . .
(b) The terms of this sequence are alternating between −1 and 1 so that the positive terms are in the even
places. Since cos πn = 1 for even n and cos πn = −1 for odd n, we have an = cos πn, n = 1, 2, . . . .
(c) The terms an are 1 for odd n and −1 for even n. Hence, an = (−1)n+1 , n = 1, 2, . . .
(d) The numerator of each term is n!, and the denominator is 2n ; hence, a = n! , n = 1, 2, 3, . . . .
n 2n

1
2. Let an = for n = 1, 2, 3, . . . . Write out the first three terms of the following sequences.
2n − 1
(a) bn = an+1 (b) cn = an+3
(c) dn = a2n (d) en = 2an − an+1
solution
(a) The first three terms of {bn } are:
1 1 1 1 1 1
b1 = a2 = = , b2 = a3 = = , b3 = a4 = =
2·2−1 3 2·3−1 5 2·4−1 7
(b) The first three terms of {cn } are:
1 1 1 1 1 1
c1 = a4 = = , c2 = a5 = = , c3 = a6 = =
2·4−1 7 2·5−1 9 2 · 6 − 1 11
(c) Note
1 1 1 1 1
a1 = = 1, a2 = = , a3 = =
2·1−1 2·2−1 3 2·3−1 5
Thus,
 2  2
1 1 1 1
d1 = a21 = 12 = 1, d2 = a22 = = , d3 = a23 = =
3 9 5 25

(d) The first three terms of {en } are:

e1 = 2a1 − a2 , e2 = 2a2 − a3 , e3 = 2a3 − a4

Thus, we must compute a1 , a2 , a3 and a4 . We set n = 1, 2, 3 and 4 in the formula for an to obtain:

1 1 1 1 1 1 1
a1 = = 1, a2 = = , a3 = = , a4 = =
2·1−1 2·2−1 3 2·3−1 5 2·4−1 7
Therefore,
1 5 1 1 7 1 1 9
e1 = 2 · 1 − = , e2 = 2 · − = , e3 = 2 · − =
3 3 3 5 15 5 7 35
 S E C T I O N 10.1 Sequences (LT SECTION 11.1) 3

In Exercises 3–12, calculate the first four terms of the sequence, starting with n = 1.
3n
3. cn =
n!
solution Setting n = 1, 2, 3, 4 in the formula for cn gives

31 3 32 9
c1 = = = 3, c2 = = ,
1! 1 2! 2
33 27 9 34 81 27
c3 = = = , c4 = = =
3! 6 2 4! 24 8

(2n − 1)!
4. bn =
n!
solution Setting n = 1, 2, 3, 4 in the formula for bn gives
(2 · 1 − 1)! 1 (2 · 2 − 1) 6
b1 = = = 1, b2 = = = 3,
1! 1 2! 2
(2 · 3 − 1)! 120 (2 · 4 − 1) 5040
b3 = = = 20, b4 = = = 210
3! 6 4! 24

5. a1 = 2, an+1 = 2a2n − 3
solution For n = 1, 2, 3 we have:

a2 = a1+1 = 2a21 − 3 = 2 · 22 − 3 = 5
a3 = a2+1 = 2a22 − 3 = 2 · 52 − 3 = 47
a4 = a3+1 = 2a23 − 3 = 2 · 472 − 3 = 4415
The first four terms of {an } are 2, 5, 47, 4415.
1
6. b1 = 1, bn = bn−1 +
bn−1
solution For n = 2, 3, 4 we have
1 1
b2 = b1 + =1+ =2
b1 1
1 1 5
b3 = b2 + =2+ =
b2 2 2
1 5 2 29
b4 = b3 + = + =
b2 2 5 10
5 29
The first four terms of {bn } are 1, 2, , .
2 10
7. bn = 5 + cos πn
solution For n = 1, 2, 3, 4 we have
b1 = 5 + cos π = 4
b2 = 5 + cos 2π = 6
b3 = 5 + cos 3π = 4
b4 = 5 + cos 4π = 6
The first four terms of {bn } are 4, 6, 4, 6.
8. cn = (−1)2n+1
solution for n = 1, 2, 3, 4 we have

c1 = (−1)2·1+1 = (−1)3 = −1
c2 = (−1)2·2+1 = (−1)5 = −1
4 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

c3 = (−1)2·3+1 = (−1)7 = −1
c4 = (−1)2·4+1 = (−1)9 = −1

The first four terms of {cn } are −1, −1, −1, −1.
1 1 1
9. cn = 1 + + + ··· +
2 3 n
solution

c1 = 1
1 3
c2 = 1 + =
2 2
1 1 3 1 11
c3 = 1 + + = + =
2 3 2 3 6
1 1 1 11 1 25
c4 = 1 + + + = + =
2 3 4 6 4 12

1 1 1
10. wn = 1 + + + ··· + 2
22 32 n
solution

w1 = 1
1 5
w2 = 1 + =
4 4
1 1 5 1 49
w3 = 1 + + = + =
4 9 4 9 36
1 1 1 49 1 205
w4 = 1 + + + = + =
4 9 16 36 16 144

11. b1 = 2, b2 = 3, bn = 2bn−1 + bn−2

solution We need to find b3 and b4 . Setting n = 3 and n = 4 and using the given values for b1 and b2 we
obtain:

b3 = 2b3−1 + b3−2 = 2b2 + b1 = 2 · 3 + 2 = 8

b4 = 2b4−1 + b4−2 = 2b3 + b2 = 2 · 8 + 3 = 19

The first four terms of the sequence {bn } are 2, 3, 8, 19.

Fn+1
12. an = where Fn is the nth Fibonacci number.
Fn
solution The first five Fibonacci numbers are F1 = 1, F2 = 1, F3 = 2, F4 = 3, F5 = 5, so

F2 1
a1 = = =1
F1 1
F3 2
a2 = = =2
F2 1
F4 3
a3 = =
F3 2
F5 5
a4 = =
F4 3

13. Find a formula for the nth term of each sequence.

1 −1 1 2 3 4
(a) , , , ... (b) , , , . . .
1 8 27 6 7 8
 S E C T I O N 10.1 Sequences (LT SECTION 11.1) 5

solution
(a) The denominators are the third powers of the positive integers starting with n = 1. Also, the sign of the
terms is alternating with the sign of the first term being positive. Thus,
1 (−1)1+1 1 (−1)2+1 1 (−1)3+1
a1 = 3
= ; a2 = − = ; a3 = =
1 13 23 23 33 33
This rule leads to the following formula for the nth term:
(−1)n+1
an =
n3
(b) Assuming a starting index of n = 1, we see that each numerator is one more than the index and the
denominator is four more than the numerator. Thus, the general term an is
n+1
an =
n+5
14. Suppose that lim an = 4 and lim bn = 7. Determine:
n→∞ n→∞
(a) lim (an + bn ) (b) lim a3n
n→∞ n→∞
(c) lim cos(πbn ) (d) lim (a2n − 2an bn )
n→∞ n→∞

solution
(a) By the Limit Laws for Sequences, we find
lim (an + bn ) = lim an + lim bn = 4 + 7 = 11
n→∞ n→∞ n→∞

(b) By the Limit Laws for Sequences, we find

       3
lim a3n = lim (an · an · an ) = lim an · lim an · lim an = lim an = 43 = 64
n→∞ n→∞ n→∞ n→∞ n→∞ n→∞

(c) By Theorem 4, we can “bring the limit inside the function”:

lim cos(πbn ) = cos(π lim bn ) = cos(7π) = −1.
n→∞ n→∞

(d) By the Limit Laws of Sequences, we find

  2   
lim a2n − 2an bn = lim a2n − lim 2an bn = lim an − 2 lim an lim bn = 42 − 2 · 4 · 7 = −40
n→∞ n→∞ n→∞ n→∞ n→∞ n→∞

In Exercises 15–28, use Theorem 1 to determine the limit of the sequence or state that the sequence diverges.
15. an = 5 − 2n
solution We have an = f (n) where f (x) = 5 − 2x; thus,
lim an = lim f (x) = lim (5 − 2x) = −∞
n→∞ x→∞ x→∞

The sequence diverges.

4
16. an = 20 − 2
n
4
solution We have an = f (n) where f (x) = 20 − ; thus,
x2
   
4 4
lim 20 − 2 = lim 20 − 2 = 20 − 0 = 20
n→∞ n x→∞ x

5n − 1
17. bn =
12n + 9
5x − 1
solution We have bn = f (n) where f (x) = ; thus,
12x + 9
5n − 1 5x − 1 5
lim = lim =
n→∞ 12n + 9 x→∞ 12x + 9 12
6 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

4 + n − 3n2
18. an =
4n2 + 1
4 + x − 3x2
solution We have an = f (n) where f (x) = ; thus,
4x2 + 1
4 + n − 3n2 4 + x − 3x2 3
lim = lim =−
n→∞ 4n + 1
2 x→∞ 4x + 1
2 4
 −n
1
19. an =
2

1 −n
solution We have an = f (n) where f (x) = 2 = 2n ; thus,
lim an = lim f (x) = lim 2 x = ∞
n→∞ x→∞ x→∞

The sequence diverges.

 n
1
20. zn =
3
 x
1
solution We have zn = f (n) where f (x) = ; thus,
3
 n  x
1 1
lim = lim =0
n→∞ 3 x→∞ 3

21. cn = 9n
solution We have cn = f (n) where f (x) = 9 x ; thus,
lim 9n = lim 9 x = ∞
n→∞ x→∞
n
Thus, the sequence 9 diverges.
22. zn = 10−1/n
solution We have zn = f (n) where f (x) = (0.1)−1/x ; thus

lim (0.1)−1/n = lim (0.1)−1/x = (0.1)limx→∞ (−1/x) = (0.1)0 = 1

n→∞ x→∞

n
23. an = √
n +1
2
x
solution We have an = f (n) where f (x) = √ ; thus,
x2 +1
x
n x 1 1 1
lim √ = lim √ = lim √
x
= lim = lim = √ =1
n→∞ n2 + 1 x→∞ x2 + 1 x→∞ x2 +1
x
x→∞ x2 +1 x→∞
1+ 1 1+0
x2 x2

n
24. an = √
n +1
3
x
solution We have an = f (n) where f (x) = √ ; thus,
x3 +1
x √1
n x 3/2 x 0 0
lim √ = lim √ = lim √x 3 = lim = √ = =0
n→∞ n3 +1 x→∞ x3 +1 x→∞ x +1 x→∞
1+ 1 1+0 1
x3/2 x3

 
12n + 2
25. an = ln
−9 + 4n
 
12x + 2
solution We have an = f (n) where f (x) = ln ; thus,
−9 + 4x
     
12n + 2 12x + 2 12x + 2
lim ln = lim ln = ln lim = ln 3
n→∞ −9 + 4n x→∞ −9 + 4x x→∞ −9 + 4x
 S E C T I O N 10.1 Sequences (LT SECTION 11.1) 7

26. rn = ln n − ln(n2 + 1)
solution We have rn = f (n) where f (x) = ln x − ln(x2 + 1); thus,
x
lim (ln n − ln(n2 + 1)) = lim (ln x − ln(x2 + 1)) = lim ln
n→∞ x→∞ x→∞ x2 + 1
But this function diverges as x → ∞, so that rn diverges as well.
n+1
27. zn = n
e
n+1
solution We have zn = f (n) where f (n) = ; thus, using L’Hôpital’s rule together with Theorems 1
en
and 2,
x+1 1
lim zn = lim f (x) = lim x
= lim x = 0
n→∞ x→∞ x→∞ e x→∞ e

1
28. yn = ne n
1
solution We have yn = f (n) where f (x) = ne n ; thus,

lim zn = lim f (x) = ∞

n→∞ x→∞

1
since lim x→∞ e x = e0 = 1. Thus the sequence diverges.

In Exercises 29–32, use Theorem 4 to determine the limit of the sequence.

1
29. an = 4+
n
solution We have

1 1
lim 4 + = lim 4 + = 4
n→∞ n x→∞ x
√
Since x is a continuous function for x > 0, Theorem 4 tells us that

1 1 √
lim 4+ = lim 4 + = 4=2
n→∞ n n→∞ n

30. an = e4n/(3n+9)
solution We have

4n 4
lim =
n→∞ 3n + 9 3
Since e x is continuous for all x, Theorem 4 tells us that

lim e4n/(3n+9) = elimn→∞ 4n/(3n+9) = e4/3

n→∞

 
−1 n3
31. an = cos
2n3 + 1
solution We have

n3 1
lim =
n→∞ 2n3 +1 2
Since cos−1 (x) is continuous for all x, Theorem 4 tells us that
   
−1 n3 −1 n3 π
lim cos = cos lim 3 = cos−1 (1/2) =
n→∞ 2n + 1
3 n→∞ 2n + 1 3
8 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

32. an = tan−1 (e−n )

solution We have

lim = e−n lim e−x = 0

n→∞ x→∞

Since tan−1 (x) is continuous for all x, Theorem 4 tells us that

 
lim tan−1 (e−n ) = tan−1 lim e−n = tan−1 (0) = 0
n→∞ n→∞

Fn+1
In Exercises 33–34 let an = , where {Fn } is the Fibonacci sequence. The sequence {an } has a limit. We
Fn
do not prove this fact, but investigate the value of the limit in these exercises.

33. Estimate lim an to five decimal places by computing an for sufficiently large n.
n→∞

solution For example,

a9 ≈ 1.61765, a99 ≈ a999 ≈ 1.61803

34. Denote the limit of {an } by L. Given that the limit exists, we can determine L as follows:
(a) Show that an+1 = 1 + a1n .
(b) Given that {an } converges to L, it follows that {an+1 } also converges to L (see Exercise 85). Show that
L2 − L − 1 = 0 and solve this equation to determine L. (The value of L is known as the golden ratio. It arises
in many different situations in mathematics.)
solution
(a) We have
1 1 Fn Fn+1 + Fn Fn+2
1+ =1+ =1+ = = = an+1
an Fn+1 /Fn Fn+1 Fn+1 Fn+1
(b) Taking the limit of both sides in part (a) gives
1
L=1+ , or L2 = L + 1, so that L2 − L − 1 = 0
L
By the quadratic formula,
√
1± 5
L=
2
Since the limit is clearly positive, we take the plus sign, giving
√
1+ 5
L=
2

n
35. Let an = . Find a number M such that:
n+1
(a) |an − 1| ≤ 0.001 for n ≥ M.
(b) |an − 1| ≤ 0.00001 for n ≥ M.
Then use the limit definition to prove that lim an = 1.
n→∞

solution
(a) We have
     
n − (n + 1)   −1 
|an − 1| =  − 1 = 
n 1
 =  =
n+1 n+1 n + 1 n + 1

Therefore |an − 1| ≤ 0.001 provided 1

n+1 ≤ 0.001, that is, n ≥ 999. It follows that we can take M = 999.
 S E C T I O N 10.1 Sequences (LT SECTION 11.1) 9

(b) By part (a), |an − 1| ≤ 0.00001 provided n+1

1
≤ 0.00001, that is, n ≥ 99999. It follows that we can take
M = 99999.
We now prove formally that lim an = 1. Using part (a), we know that
n→∞

1
|an − 1| = <
n+1
provided n > 1
 − 1. Thus, Let  > 0 and take M = 1
 − 1. Then, for n > M, we have
1 1
|an − 1| = < =
n+1 M+1
1 n
36. Let bn = 3 .
(a) Find a value of M such that |bn | ≤ 10−5 for n ≥ M.
(b) Use the limit definition to prove that lim bn = 0.
n→∞

solution
(a) Solving 1 n
3 ≤ 10−5 for n, we find
ln 10
n ≥ 5 log3 10 = 5 ≈ 10.48
ln 3
It follows that we can take M = 10.5.
(b) We see that
 n 
 1  1
 − 0 = n < 
3  3
provided
1
n > log3

Thus, let  > 0 and take M = log3 1 . Then, for n > M, we have
 n 
 1  1 1
 − 0 = n < M = 
3  3 3

37. Use the limit definition to prove that lim n−2 = 0.

n→∞

solution We see that

 
− 0| =  2  = 2 < 
−2 1 1
|n
n n
provided
1
n> √

Thus, let  > 0 and take M = √1 .

Then, for n > M, we have
 
|n−2 − 0| =  2  = 2 < 2 = 
1 1 1
n n M
n
38. Use the limit definition to prove that lim = 1.
n→∞ n + n−1
solution Since

n n2 n2
−1
= −1
= 2
n+n n(n + n ) n + 1
we see that
 2   
 n   −1 
− 1 =  2
1
 2  = 2 <
n +1  n +1 n +1
10 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

provided

1
n> −1


1
So choose  > 0, and let M = − 1. Then, for n > M, we have

   
 n −  =  −1  = 1 < 1
 n + n−1 1   n2 + 1  n2 + 1 ( 1 − 1) + 1 = 


In Exercises 39–66, use the appropriate limit laws and theorems to determine the limit of the sequence or
show that it diverges.
 n
1
39. an = 10 + −
9
solution By the Limit Laws for Sequences we have:
  n   n  n
1 1 1
lim 10 + − = lim 10 + lim − = 10 + lim −
n→∞ 9 n→∞ n→∞ 9 n→∞ 9
Now for all n,
 n  n  n
1 1 1
− ≤ − ≤
9 9 9
We will use the Squeeze Theorem. The limit of the right-hand side of the above inequality is zero since it is
a geometric series with ratio less than 1. The limit of the left-hand side is also zero, since
  n   n
1 1
lim − = − lim =0
n→∞ 9 n→∞ 9

Therefore
  n   n  n
1 1 1
10 = 10 + lim − ≤ 10 + lim − ≤ 10 + lim = 10
n→∞ 9 n→∞ 9 n→∞ 9

so by the Squeeze Theorem,

  n 
1
lim 10 + − = 10
n→∞ 9
√ √
40. dn = n+3− n
√ √
solution We multiply and divide dn by the conjugate expression n + 3 + n and use the identity (a −
b)(a + b) = a2 − b2 to obtain:
√ √ √ √
n+3− n n+3+ n (n + 3) − n 3
dn = √ √ = √ √ = √ √
n+3+ n n+3+ n n+3+ n
Thus,
3 3
lim dn = lim √ √ = x→∞
lim √ √ =0
n→∞ n→∞ n+3+ n x+3+ x

41. cn = 1.01n
solution Since cn = f (n) where f (x) = 1.01 x , we have

lim 1.01n = lim 1.01 x = ∞

n→∞ x→∞

so that the sequence diverges.

 S E C T I O N 10.1 Sequences (LT SECTION 11.1) 11

2
42. bn = e1−n
2
solution Since bn = f (n) where f (x) = e1−x , we have
2 2 e
lim e1−n = lim e1−x = lim =0
n→∞ x→∞ x→∞ e x2

43. an = 21/n
solution Because 2 x is a continuous function,

lim 21/n = lim 21/x = 2limx→∞ (1/x) = 20 = 1

n→∞ x→∞

44. bn = n1/n
solution Let an = n1/n . Take the natural logarithm of both sides of this expression to obtain
ln n
ln an = ln n1/n =
n
Thus,
ln n ln x 1
lim (ln an ) = lim = lim = lim = 0
n→∞ n→∞ n x→∞ x x→∞ x

Because f (x) = e x is a continuous function, it follows that

lim an = lim eln an = elimn→∞ (ln an ) = e0 = 1
n→∞ n→∞

That is,
lim n1/n = 1
n→∞

9n
45. cn =
n!
solution For n ≥ 9, write
9n 9 9 9 9 9 9 9
cn = = · ··· · · ··· ·
n! 
1 2 n−1 n
9 
10 11
call this C Each factor is less than 1

Then clearly
9n 9
0≤ ≤C
n! n
since each factor after the first nine is < 1. The squeeze theorem tells us that
9n 9 9
lim 0 ≤ lim ≤ lim C = C lim = C · 0 = 0
n→∞ n→∞ n! n→∞ n n→∞ n

so that limn→∞ cn = 0 as well.

82n
46. an =
n!
solution Note that

82n 64n
an = =
n! n!
Now apply the same method as in the Exercise 45. For n ≥ 64, write
64n 64 64 64 64 64 64 64
cn = = · ··· · · ··· ·
n! 1 2 64 65 66 n −1 n
 
call this C Each factor is less than 1

Then clearly
64n 64
0≤ ≤C
n! n
12 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

since each factor after the first 64 is < 1. The squeeze theorem tells us that
64n 64 64
lim 0 ≤ lim ≤ lim C = C lim =C·0=0
n→∞ n→∞ n! n→∞ n n→∞ n

so that limn→∞ an = 0 as well.

3n2 + n + 2
47. an =
2n2 − 3
solution

3n2 + n + 2 3x2 + x + 2 3
lim = lim =
n→∞ 2n2 − 3 x→∞ 2x2 − 3 2
√
n
48. an = √
n+4
solution
√
√ √ √
x
n x x 1 1
lim √ = lim √ = lim √ = lim = =1
n→∞ n + 4 x→∞ x + 4 x→∞ √
x
+ √4 x→∞ 1 + √4x 1+0
x x

cos n
49. an =
n
solution Since −1 ≤ cos n ≤ 1 the following holds:
1 cos n 1
− ≤ ≤
n n n
We now apply the Squeeze Theorem for Sequences and the limits
1 1
lim − = lim = 0
n→∞ n n→∞ n
to conclude that lim cos n
n = 0.
n→∞
(−1)n
50. cn = √
n
solution Clearly
1 (−1)n 1
−√ ≤ √ ≤ √
n n n
Since
−1 1
lim √ = lim √ = 0
n→∞ n n→∞ n
(−1)n
the Squeeze Theorem tells us that lim √
n
= 0.
n→∞
51. dn = ln 5n − ln n!
solution Note that
5n
dn = ln
n!
so that
5n 5n
edn = so lim edn = lim =0
n! n→∞ n→∞ n!

by the method of Exercise 45. If dn converged, we could, since f (x) = e x is continuous, then write
lim edn = elimn→∞ dn = 0
n→∞

which is impossible. Thus {dn } diverges.

 S E C T I O N 10.1 Sequences (LT SECTION 11.1) 13

52. dn = ln(n2 + 4) − ln(n2 − 1)

solution Note that

n2 + 4
dn = ln
n2 − 1
so that
n2 + 4 n2 + 4
edn = so lim edn = lim =1
n2 − 1 n→∞ n→∞ n2 − 1

by the method of Exercise 45. Since f (x) = e x is continuous, we have

lim edn = 1
n→∞

so that lim dn = ln 1 = 0.
n→∞
 1/3
4
53. an = 2 + 2
n

4 1/3
solution Let an = 2 + n2
. Taking the natural logarithm of both sides of this expression yields
 1/3  
4 1 4
ln an = ln 2 + 2 = ln 2 + 2
n 3 n
Thus,
 1/3     
1 4 1 4 1 4
lim ln an = lim ln 2 + 2 = lim ln 2 + 2 = ln lim 2 + 2
n→∞ n→∞ 3 n 3 x→∞ x 3 x→∞ x
1 1
= ln (2 + 0) = ln 2 = ln 21/3
3 3
Because f (x) = e x is a continuous function, it follows that
1/3
lim an = lim eln an = elimn→∞ (ln an ) = eln 2 = 21/3
n→∞ n→∞

 
2
54. bn = tan−1 1 −
n
solution Because f (x) = tan−1 x is a continuous function, it follows that
    
2 2 π
lim an = lim tan−1 1 − = tan−1 lim 1 − = tan−1 1 =
n→∞ x→∞ x x→∞ x 4
 
2n + 1
55. cn = ln
3n + 4
solution Because f (x) = ln x is a continuous function, it follows that
   
2x + 1 2x + 1 2
lim cn = lim ln = ln lim = ln
n→∞ x→∞ 3x + 4 x→∞ 3x + 4 3

n
56. cn =
n + n1/n
solution We rewrite n
n+n1/n
as follows:
n
n 1
= n
=
n + n1/n n
+ n1/n
1+ n1/n
n n n

Now,
n1/n 1
= n n −1 = 1−1/n
1

n n
14 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

and
n1/n 1 1
lim = lim 1−1/n = lim 1−1/x = 0
n→∞ n n→∞ n x→∞ x

Thus,
n 1 lim x→∞ 1 lim x→∞ 1 1
lim = lim =  = = =1
n→∞ n + n1/n x→∞ 1 + x 1/x
lim x→∞ 1 + x x 1/x
lim x→∞ 1 + lim x→∞ x1/x 1+0
x x

en
57. yn =
2n
n
 n  n
solution 2en = 2e and 2e > 1. By the Limit of Geometric Sequences,we conclude that limn→∞ 2e = ∞.
Thus, the given sequence diverges.
n
58. an = n
2
solution
d
n x dx (x) 1 1 1 1
lim = lim x = lim = lim = lim = ·0=0
n→∞ 2n x→∞ 2 x→∞ d
(2 x ) x→∞ (ln 2) 2 x ln 2 x→∞ 2 x ln 2
dx

en + (−3)n
59. yn =
5n
solution
 e n  n
en + (−3)n −3
lim n
= lim + lim
n→∞ 5 n→∞ 5 n→∞ 5

assuming both limits on the right-hand side exist. But by the Limit of Geometric Sequences, since
−3 e
−1 < <0< <1
5 5
both limits on the right-hand side are 0, so that yn converges to 0.
(−1)n n3 + 2−n
60. bn =
3n3 + 4−n
solution Assuming both limits on the right-hand side exist, we have
(−1)n n3 + 2−n (−1)n n3 2−n
lim −n
= lim 3 −n
+ lim 3
n→∞ 3n + 4
3 n→∞ 3n + 4 n→∞ 3n + 4−n
For the first limit, let us consider instead the limit of its reciprocal:
3n3 + 4−n n 3n
3
n4
−n
lim (−1)n = lim (−1) + lim (−1)
n→∞ n3 n→∞ n3 n→∞ n3
1
= lim (−1)n · 3 + lim (−1)n n 3
n→∞ n→∞ 4n
= lim ((−1)n · 3) + 0
n→∞

so that one limit on the right-hand side exists and the other does not; thus the left-hand side diverges as well.
π
61. an = n sin
n
solution By the Theorem on Sequences Defined by a Function, we have
π π
lim n sin = lim x sin
n→∞ n x→∞ x
Now,
 
π sin π cos πx − xπ2  π
lim x sin = lim 1 x
= lim = lim π cos
x→∞ x x→∞ x x→∞ − x12 x→∞ x
π
= π lim cos = π cos 0 = π · 1 = π
x→∞ x
 S E C T I O N 10.1 Sequences (LT SECTION 11.1) 15

Thus,
π
lim n sin =π
n→∞ n

n!
62. bn =
πn
4n n!
solution By the method of Exercise 45, we can see that limn→∞ = 0 so that cn = n diverges. But π < 4
n! 4
so that cn < bn and thus bn diverges as well.
3 − 4n
63. bn =
2 + 7 · 4n
solution Divide the numerator and denominator by 4n to obtain
4n
4n − 4n −1
3 3
3 − 4n 4n
an = = =
2 + 7 · 4n 4n + 4n
2 7·4n 2
4n +7

Thus,

3
x −1 lim x→∞ 3
4x −1 3 lim x→∞ 1
4x − lim x→∞ 1 3·0−1 1
lim an = lim 42 =  = = =−
n→∞ x→∞ x
4 +7 lim x→∞ 42x +7 2 lim x→∞ 41x − lim x→∞ 7 2·0+7 7

3 − 4n
64. an =
2 + 7 · 3n
solution Divide the numerator and denominator by 3n to obtain

4 n
3 − 4n
3 4n
3n − 3n
3
3n − 3
an = = =
2 + 7 · 3n 3n + 3n
2 7·3n 2
3n +7

We examine the limits of the numerator and the denominator:

  n   n  n
3 4 1 4
lim − = 3 lim − 3 lim = 3 · 0 − ∞ = −∞
n→∞ 3n 3 n→∞ 3 n→∞ 3

whereas
   n
2 2 1
lim + 7 = lim + lim 7 = 2 lim + lim 7 = 2 · 0 + 7 = 7
n→∞ 3n n→∞ 3n n→∞ n→∞ 3 n→∞

Thus, lim an = −∞; that is, the sequence diverges.

n→∞
 n
1
65. an = 1 +
n
solution Taking the natural logarithm of both sides of this expression yields
 n   ln 1 + 1
1 1 n
ln an = ln 1 + = n ln 1 + = 1
n n n

Thus,
   
ln 1 + 1 d
ln 1 + 1 1
1+ 1x
· − x12 1 1
x dx x
lim (ln an ) = lim = lim  = lim = lim = =1
n→∞ x→∞ 1
x
x→∞ d 1 x→∞ − x12 x→∞ 1+ 1
x
1+0
dx x

Because f (x) = e x is a continuous function, it follows that

lim an = lim eln an = elimn→∞ (ln an ) = e1 = e

n→∞ n→∞
16 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)
 n
1
66. an = 1 + 2
n
solution Taking the natural logarithm of both sides of this expression yields
 n   ln 1 + 1
1 1 n2
ln an = ln 1 + 2 = n ln 1 + 2 = 1
n n n

Thus,

ln(1 + x−2 )
d
dx ln(1 + x−2 )
lim (ln an ) = lim = lim
n→∞ x→∞ x−1 x→∞ d −1
dx (x )
1
1+x−2
(−2x−3 ) 2x−1
2
0
= lim = lim = lim x
= =0
x→∞ −x−2 x→∞ 1 + x−2 x→∞ 1 + 1
x2
1+0
Because f (x) = e x is a continuous function, it follows that
lim an = lim eln an = elimn→∞ (ln an ) = e0 = 1
n→∞ n→∞

In Exercises 67–70, find the limit of the sequence using L’Hôpital’s Rule.
(ln n)2
67. an =
n
solution d 2 2 ln x
(ln n)2 (ln x)2 dx (ln x) 2 ln x
lim = lim = lim d
= lim x
= lim
n→∞ n x→∞ x x→∞
dx x
x→∞ 1 x→∞ x
d 2
dx 2 ln x 2
= lim d
= lim x
= lim =0
x→∞
dx x
x→∞ 1 x→∞ x
 
√ 1
68. bn = n ln 1 +
n
solution  
    ln 1 + 1x d
ln 1 + 1
√ 1 √ 1 dx x
lim n ln 1 + = lim x ln 1 + = lim = lim
n→∞ n x→∞ x x→∞ x−1/2 x→∞ d −1/2
dx x

1
1+ 1x
· −1 x2 2
= lim −1 = lim √  =0
−3/2
x→∞
2 x x→∞ x 1 + 1x
√
69. cn = n n2 + 1 − n
solution √ √
√ √ x x2 + 1 − x x2 + 1 + x
lim n n2 + 1 − n = lim x x2 + 1 − x = lim √
n→∞ x→∞ x→∞ x2 + 1 + x
d
x dx x 1
= lim √ = lim √ = lim
x→∞ x2 +1+x x→∞ d x2 +1+x x→∞ 1+ √x
dx x2 +1

1 1 1
= lim = lim =
x→∞ x→∞
1+ 1 2
1+ x2
x2 +1 1+(1/x2 )

√3
70. dn = n2 n3 + 1 − n
solution We rewrite an as follows:
⎛   ⎞ ⎛ ⎞
 √3 ⎜⎜⎜ ⎟⎟⎟ ⎜ ⎟⎟
2⎜
n + 1 − n = n ⎜⎜⎝ n 1 + 3 − n⎟⎟⎠ = n ⎜⎜⎜⎝n
2⎜ 1 ⎟ 1 + 3 − n⎟⎟⎟⎠
⎜ ⎟ 3 1
dn = n2 3 3 3
n n
⎛ ⎞  1/3
⎜⎜ 3 ⎟⎟⎟ (1 + n−3 ) − 1
3⎜ ⎜ 1 ⎟
= n ⎝⎜ 1 + 3 − 1⎠⎟ =
n n−3
 S E C T I O N 10.1 Sequences (LT SECTION 11.1) 17

Thus,
 1/3

(1 + x−3 )
1/3
−1
d
dx (1 + x−3 ) −1
lim dn = lim = lim
n→∞ x→∞ x−3 x→∞ d −3
dx [x ]
−2/3
1
3 (1 + x−3 ) (−3x−4 ) 1 −2/3 1 1
= lim = lim (1 + x−3 ) = lim  =
x→∞ −3x−4 x→∞ 3 x→∞
3 1 + x13
2/3 3

In Exercises 71–74, use the Squeeze Theorem to evaluate lim an by verifying the given inequality.
n→∞

1 1 1
71. an = √ , √ ≤ an ≤ √
n4 + n8 2n4 2n2
solution For all n > 1 we have n4 < n8 , so the quotient √ 1 is smaller than √ 1 and larger than √ 1 .
n4 +n8 n4 +n4 n8 +n8
That is,

1 1 1
an < √ = √ = √ ; and
n4 + n4
n4 · 2 2n2
11 1
an > √ = √ = √
n +n
8 8 2n 8 2n4

1 1
Now, since lim √ = lim √ = 0, the Squeeze Theorem for Sequences implies that lim an = 0.
n→∞ 2n4 n→∞ 2n2 n→∞

1 1 1
72. cn = √ + √ + ··· + √ ,
n2 +1 n2 +2 n +n
2
n n
√ ≤ cn ≤ √
n2 + n n2 + 1
solution Since each of the n terms in the sum defining cn is not smaller than √1 and not larger than
n2 +n
√1 we obtain the following inequalities:
n2 +1

1 1 1 n
cn ≥ √ + ··· + √ =n· √ = √ ;
n +n
2 n +n2
 n +n
2 n +n
2

n terms

1 1 1 n
cn ≤ √ + ··· + √ =n· √ = √
n +1
2 n +12
 n +1
2 n +1
2

n terms

Thus,
n n
√ ≤ cn ≤ √
n2 +n n2 +1
We now compute the limits of the two sequences:
n
n 1 1
lim √ = lim √
n
= lim √ = lim = 1;
n→∞ n2 +1 n→∞ n2 +1 n→∞ n2 +1 n→∞
n
√
n2 1+ 1
n2
n
n 1 1
lim √ = lim √
n
= lim √ = lim =1
n→∞ n2 +n n→∞ n2 +n n→∞ √n2 +n n→∞
n n2 1+ 1
n

By the Squeeze Theorem we conclude that:

lim cn = 1
n→∞
18 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

73. an = (2n + 3n )1/n , 3 ≤ an ≤ (2 · 3n )1/n = 21/n · 3

solution Clearly 2n + 3n ≥ 3n for all n ≥ 1. Therefore:

(2n + 3n )1/n ≥ (3n )1/n = 3

Also 2n + 3n ≤ 3n + 3n = 2 · 3n , so
(2n + 3n )1/n ≤ (2 · 3n )1/n
Thus,
3 ≤ (2n + 3n )1/n ≤ (2 · 3n )1/n
Because
lim (2 · 3n )1/n = lim 21/n · 3 = 3 lim 21/n = 3 · 1 = 3
n→∞ n→∞ n→∞

and limn→∞ 3 = 3, the Squeeze Theorem for Sequences guarantees

lim (2n + 3n )1/n = 3
n→∞

74. an = (n + 10n )1/n , 10 ≤ an ≤ (2 · 10n )1/n

solution Clearly
10n ≤ n + 10n ≤ 10n + 10n = 2 · 10n
for all n ≥ 0. Thus
10 ≤ (n + 10n )1/n ≤ (2 · 10n )1/n
Now,
lim (2 · 10n )1/n = lim 21/n · 10 = 10 lim 21/n = 10 · 1 = 10
n→∞ n→∞ n→∞

and limn→∞ 10 = 10, so that the Squeeze Theorem for Sequences tells us that
lim (n + 10n )1/n = 10
n→∞

75. Which of the following statements is equivalent to the assertion lim an = L? Explain.
n→∞
(a) For every  > 0, the interval (L − , L + ) contains at least one element of the sequence {an }.
(b) For every  > 0, the interval (L − , L + ) contains all but at most finitely many elements of the sequence
{an }.
solution Statement (b) is equivalent to Definition 1 of the limit, since the assertion “|an − L| <  for all
n > M” means that L −  < an < L +  for all n > M; that is, the interval (L − , L + ) contains all the
elements an except (maybe) the finite number of elements a1 , a2 , . . . , a M .
Statement (a) is not equivalent to the assertion lim an = L. We show this, by considering the following
n→∞
sequence:
⎧1
⎪
⎪
⎪
⎪
⎪
⎨n
for odd n
an = ⎪
⎪
⎪
⎪
⎪ 1
⎩1 + for even n
n
Clearly for every  > 0, the interval (−, ) = (L − , L + ) for L = 0 contains at least one element of {an },
but the sequence diverges (rather than converges to L = 0). The terms in the odd places converge to 0 and the
terms in the even places converge to 1. Hence, an does not approach one limit.
1
76. Show that an = is decreasing.
2n + 1
solution Let f (x) = 2x+11
. Then
1 −2 1
f  (x) = − ·2= <0 for x  −
(2x + 1)2 (2x + 1)2 2
Since f  (x) < 0 for x  − 12 , f is decreasing on the interval x > − 12 . It follows that an = f (n) is also decreas-
ing.
 S E C T I O N 10.1 Sequences (LT SECTION 11.1) 19

3n2
77. Show that an = is increasing. Find an upper bound.
n2 + 2
3x2
solution Let f (x) = x2 +2
. Then

6x(x2 + 2) − 3x2 · 2x 12x

f  (x) = =
(x2 + 2)2 (x2 + 2)2
f  (x) > 0 for x > 0, hence f is increasing on this interval. It follows that an = f (n) is also increasing. We
now show that M = 3 is an upper bound for an , by writing:

3n2 3n2 + 6 3(n2 + 2)

an = ≤ 2 = 2 =3
n2 +2 n +2 n +2
That is, an ≤ 3 for all n.
√3
78. Show that an = n + 1 − n is decreasing.
√3
solution Let f (x) = x + 1 − x. Then

d  1
f  (x) = (x + 1)1/3 − x = (x + 1)−2/3 − 1
dx 3
For x ≥ 1,
1 1
(x + 1)−2/3 − 1 ≤ 2−2/3 − 1 < 0
3 3
We conclude that f is decreasing on the interval x ≥ 1. It follows that an = f (n) is also decreasing.
79. Give an example of a divergent sequence {an } such that lim |an | converges.
n→∞

solution Let an = (−1)n . The sequence {an } diverges because the terms alternate between +1 and −1;
however, the sequence {|an |} converges because it is a constant sequence, all of whose terms are equal to 1.
80. Give an example of divergent sequences {an } and {bn } such that {an + bn } converges.
solution Let an = 2n and bn = −2n . Then {an } and {bn } are divergent geometric sequences. However, since
an + bn = 2n − 2n = 0, the sequence {an + bn } is the constant sequence with all the terms equal zero, so it
converges to zero.
81. Using the limit definition, prove that if {an } converges and {bn } diverges, then {an + bn } diverges.
solution We will prove this result by contradiction. Suppose limn→∞ an = L1 and that {an + bn } converges
to a limit L2 . Now, let  > 0. Because {an } converges to L1 and {an + bn } converges to L2 , it follows that there
exist numbers M1 and M2 such that:

|an − L1 | < for all n > M1 ,
2

| (an + bn ) − L2 | < for all n > M2
2
Thus, for n > M = max{M1 , M2 },
 
|an − L1 | < and | (an + bn ) − L2 | <
2 2
By the triangle inequality,

|bn − (L2 − L1 )| = |an + bn − an − (L2 − L1 )| = |(−an + L1 ) + (an + bn − L2 )|

≤ |L1 − an | + |an + bn − L2 |

Thus, for n > M,

 
|bn − (L2 − L1 ) | < + =
2 2
that is, {bn } converges to L2 − L1 , in contradiction to the given data. Thus, {an + bn } must diverge.
20 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

82. Use the limit definition to prove that if {an } is a convergent sequence of integers with limit L, then there
exists a number M such that an = L for all n ≥ M.
solution Suppose {an } converges to L, and let  = 12 . Then, there exists a number M such that

1
|an − L| <
2
for all n ≥ M. In other words, for all n ≥ M,
1 1
L− < an < L +
2 2
However, we are given that {an } is a sequence of integers. Thus, it must be that an = L for all n ≥ M.
83. Theorem 1 states that if lim f (x) = L, then the sequence an = f (n) converges and lim an = L. Show that
x→∞ n→∞
the converse is false. In other words, find a function f such that an = f (n) converges but lim f (x) does not
x→∞
exist.
solution Let f (x) = sin πx and an = sin πn. Then an = f (n). Since sin πx is oscillating between −1
and 1 the limit lim f (x) does not exist. However, the sequence {an } is the constant sequence in which
x→∞
an = sin πn = 0 for all n, hence it converges to zero.
84. Use the limit definition to prove that the limit does not change if a finite number of terms are added or
removed from a convergent sequence.
solution Suppose that {an } is a sequence such that limn→∞ an = L. For every  > 0, there is a number
M such that |an − L| <  for all n > M. That is, the inequality |an − L| <  holds for all the terms of {an }
except possibly a finite number of terms. If we add a finite number of terms, these terms may not satisfy the
inequality |an − L| < , but there are still only a finite number of terms that do not satisfy this inequality. By
removing terms from the sequence, the number of terms in the new sequence that do not satisfy |an − L| < 
are no more than in the original sequence. Hence the new sequence also converges to L.
85. Let bn = an+1 . Use the limit definition to prove that if {an } converges, then {bn } also converges and
lim an = lim bn .
n→∞ n→∞

solution Suppose {an } converges to L. Let bn = an+1 , and let  > 0. Because {an } converges to L, there exists
an M  such that |an − L| <  for n > M  . Now, let M = M  − 1. Then, whenever n > M, n + 1 > M + 1 = M  .
Thus, for n > M,

|bn − L| = |an+1 − L| < 

Hence, {bn } converges to L.

86. Let {an } be a sequence such that lim |an | exists and is nonzero. Show that lim an exists if and only if
n→∞ n→∞
there exists an integer M such that the sign of an does not change for n > M.
solution Let {an } be a sequence such that lim |an | exists and is nonzero. Suppose lim an exists and let
n→∞ n→∞
L = lim an . Note that L cannot be zero for then lim |an | would also be zero. Now, choose  < |L|. Then there
n→∞ n→∞
exists an integer M such that |an − L| < , or L −  < an < L + , for all n > M. If L < 0, then −2L < an < 0,
whereas if L > 0, then 0 < an < 2L; that is, an does not change for n > M.
Now suppose that there exists an integer M such that an does not change for n > M. If an > 0 for n > M,
then an = |an | for n > M and

lim an = lim |an |

n→∞ n∞

On the other hand, if an < 0 for n > M, then an = −|an | for n > M and

lim an = lim −|an | = − lim |an |

n→∞ n∞ n→∞

In either case, lim an exists. Thus, lim an exists if and only if there exists an integer M such that the sign of
n→∞ n→∞
an does not change for n > M.
 S E C T I O N 10.1 Sequences (LT SECTION 11.1) 21

√ √ √
87. Proceed as in Example 13 to show that the sequence 3, 3 3, 3 3 3, . . . is increasing and
bounded above by M = 3. Then prove that the limit exists and find its value.
solution This sequence is defined recursively by the formula:
 √
an+1 = 3an , a1 = 3

Consider the following inequalities:

√ √ √
a2 = 3a1 = 3 3 > 3 = a1 ⇒ a2 > a1 ;
√ √
a3 = 3a2 > 3a1 = a2 ⇒ a3 > a2 ;
√ √
a4 = 3a3 > 3a2 = a3 ⇒ a4 > a3

In general, if we assume that ak > ak−1 , then

 
ak+1 = 3ak > 3ak−1 = ak

√ induction, an+1 > an for all n; that is, the sequence

Hence, by mathematical √ {an } is increasing.
Because an+1 = 3an , it follows that an ≥ 0 for all n. Now, a1 = 3 < 3. If ak ≤ 3, then
 √
ak+1 = 3ak ≤ 3 · 3 = 3

Thus, by mathematical induction, an ≤ 3 for all n.

Since {an } is increasing and bounded, it follows by the Theorem on Bounded Monotonic Sequences that
this sequence is converging. Denote the limit by L = limn→∞ an . Using Exercise 85, it follows that
 √
L = lim an+1 = lim 3an = 3 lim an = 3L
n→∞ n→∞ n→∞
√
Thus, L2 = 3L, so L = 0 or L√ = 3. Because the sequence is increasing, we have an ≥ a1 = 3 for all n. Hence,
the limit also satisfies L ≥ 3. We conclude that the appropriate solution is L = 3; that is, lim an = 3.
n→∞

88. Let {an } be the sequence defined recursively by


a0 = 0, an+1 = 2 + an

√ √ √
Thus, a1 = 2, a2 = 2 + 2, a3 = 2 + 2 + 2, . . . .
(a) Show that if an < 2, then an+1 < 2. Conclude by induction that an < 2 for all n.
(b) Show that if an < 2, then an ≤ an+1 . Conclude by induction that {an } is increasing.
√
(c) Use (a) and (b) to conclude that L = lim an exists. Then compute L by showing that L = 2 + L.
n→∞

solution
(a) Assume an < 2. Then
 √
an+1 = 2 + an < 2 + 2 = 2

so that an+1 < 2. So by induction, an < 2 for all n and {an } is bounded above by 2.
(b) Assume an < 2. Then
 √ 
an+1 = 2 + an > an + an = 2an > a2n = an

so that an < an+1 . It follows by induction that {an } is increasing.

(c) Since {an } is increasing and bounded above, the Theorem on Bounded Monotone Sequences tells us that
L = limn→∞ an exists. We have
 √
L = lim an+1 = lim 2 + an = 2 + lim an = 2 + L
n→∞ n→∞ n→∞
√
by Exercise 85. It follows that L = 2 + L, so that L2 − L − 2 = 0. Thus L = 2 or L = −1. But all terms of
{an } are positive, so we must have L = 2.
22 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

Further Insights and Challenges

√n
89. Show that lim n! = ∞. Hint: Verify that n! ≥ (n/2)n/2 by observing that half of the factors of n! are
n→∞
greater than or equal to n/2.
 n/2
solution We show that n! ≥ n2 . For n ≥ 4 even, we have:
n n  n 
n! = 1 · · · · · · + 1 · ··· · n ≥ + 1 · · · · · n.
2 
 2 2

n n n
2 factors 2 factors 2 factors

n
Since each one of the 2 factors is greater than n2 , we have:
n  n n  n n/2
n! ≥ + 1 · ··· · n ≥ · ··· · =
2
 2 2
 2
n n
2 factors 2 factors

For n ≥ 3 odd, we have:

n−1 n+1 n+1
n! = 1 · · · · · · · ··· · n ≥ · ··· · n
2 2 2
  
n−1 n+1 n+1
2 factors 2 factors 2 factors

n+1
Since each one of the 2 factors is greater than n2 , we have:

n+1 n n  n (n+1)/2  n n/2 n  n n/2

n! ≥ · ··· · n ≥ · ··· · = = ≥
2
 2 2
 2 2 2 2
n+1 n+1
2 factors 2 factors


n n/2
In either case we have n! ≥ 2 . Thus,
√n n
n! ≥
2
n √n √n
Since lim = ∞, it follows that lim n! = ∞. Thus, the sequence an = n! diverges.
2
n→∞ n→∞
√n
n!
90. Let bn = .
n
1 k
n
(a) Show that ln bn = ln .
n k=1 n
1
(b) Show that ln bn converges to ln x dx, and conclude that
0
−1
bn → e .
solution
 1/n
n!
(a) Let bn = n . Then
1 ln (n!) − n ln n 1 ! " 1 n!
ln bn = ln (n!)1/n − ln n = ln (n!) − ln n = = ln (n!) − ln nn = ln n
n n n n n
   
1 k
n
1 1 2 3 n 1 1 2 3 n
= ln · · · ··· · = ln + ln + ln + · · · + ln = ln
n n n n n n n n n n n k=1 n

(b) By part (a) we have,

1 k
n
lim (ln bn ) = lim ln
n→∞ n→∞ n n
k=1
1 #n
Notice that n k=1 ln nk is the nth right-endpoint approximation of the integral of ln x over the interval [0, 1].
Hence,
1 k
n 1
lim ln = ln x dx
n→∞ n n 0
k=1
 S E C T I O N 10.1 Sequences (LT SECTION 11.1) 23

We compute the improper integral using integration by parts, with u = ln x and v = 1. Then u = 1x , v = x
and
1 1 1 1
ln x dx = x ln x −
1
x dx = 1 · ln 1 − lim (x ln x) − dx
0 0 0 x x→0+ 0
1
= 0 − lim (x ln x) − x = −1 − lim (x ln x)
x→0+ 0 x→0+

We compute the remaining limit using L’Hôpital’s Rule. This gives:

1
ln x
lim (x · ln x) = lim = lim x
= lim (−x) = 0
x→0+ x→0+ 1 x→0+ − 12 x→0+
x x

Thus,
1
lim ln bn = ln x dx = −1
n→∞ 0

and
lim bn = e−1
n→∞

91. Given positive numbers a1 < b1 , define two sequences recursively by

 an + bn
an+1 = an bn , bn+1 =
2
(a) Show that an ≤ bn for all n (Figure 14).
(b) Show that {an } is increasing and {bn } is decreasing.
bn − an
(c) Show that bn+1 − an+1 ≤ .
2
(d) Prove that both {an } and {bn } converge and have the same limit. This limit, denoted AGM(a1 , b1 ), is called
the arithmetic-geometric mean of a1 and b1 .
√
(e) Estimate AGM(1, 2) to three decimal places.

Geometric Arithmetic
mean mean
x
an a n+1 b n+1 bn
AGM(a 1, b 1)

FIGURE 14

solution
(a) Use induction on n. The statement is true for n = 1. Choose k > 1; then
√ √ 2 √ √ √ 2
ak−1 + bk−1  ak−1 + bk−1 − 2 ak−1 bk−1 ak−1 − 2 ak−1 bk−1 + bk−1
bk − ak = − ak−1 bk−1 = =
2 2 2
√ √ 2
ak−1 − bk−1
= ≥0
2
We conclude that bk ≥ ak , so by induction bn ≥ an for all n.
(b) By part (a), bn ≥ an for all n, so
 √
an+1 = an bn ≥ an · an = a2n = an

for all n. Hence, the sequence {an } is increasing. Moreover, since an ≤ bn for all n,
an + bn bn + bn 2bn
bn+1 = ≤ = = bn
2 2 2
for all n; that is, the sequence {bn } is decreasing.
(c) Since {an } is increasing, an+1 ≥ an . Thus,
an + bn an + bn − 2an bn − an
bn+1 − an+1 ≤ bn+1 − an = − an = =
2 2 2
24 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

(d) By part (a), an ≤ bn for all n. By part (b), {bn } is decreasing. Hence bn ≤ b1 for all n. Combining the
two inequalities we conclude that an ≤ b1 for all n. That is, the sequence {an } is increasing and bounded
(0 ≤ an ≤ b1 ). By the Theorem on Bounded Monotonic Sequences we conclude that {an } converges. Sim-
ilarly, since {an } is increasing, an ≥ a1 for all n. We combine this inequality with bn ≥ an to conclude that
bn ≥ a1 for all n. Thus, {bn } is decreasing and bounded (a1 ≤ bn ≤ b1 ); hence this sequence converges.
To show that {an } and {bn } converge to the same limit, note that
bn−1 − an−1 bn−2 − an−2 b1 − a1
bn − an ≤ ≤ ≤ · · · ≤ n−1
2 22 2
Thus,
1
lim (bn − an ) = (b1 − a1 ) lim =0
n→∞ n→∞ 2n−1
(e) We have
 an + bn √
an+1 = an bn , a1 = 1; , b1 = 2
bn+1 =
2
Computing the values of an and bn until the first three decimal digits are equal in successive terms, we obtain:
 √
a2 = a1 b1 = 1·
2 = 1.1892
√
a1 + b1 1 + 2
b2 = = = 1.2071
2 2
 √
a3 = a2 b2 = 1.1892 · 1.2071 = 1.1981
a2 + b2 1.1892 + 1.2071
b3 = = = 1.1981
2 2

a4 = a3 b3 = 1.1981
a3 + b3
b4 = = 1.1981
2
Thus,
 √
AGM 1, 2 ≈ 1.198

1 1 1 1
92. Let cn = + + + ··· + .
n n+1 n+2 2n
(a) Calculate c1 , c2 , c3 , c4 .
(b) Use a comparison of rectangles with the area under y = x−1 over the interval [n, 2n] to prove that
2n 2n
dx 1 dx 1
+ ≤ cn ≤ +
n x 2n n x n
(c) Use the Squeeze Theorem to determine lim cn .
n→∞

solution
(a)
1 3
c1 = 1 + = ;
2 2
1 1 1 13
c2 = + + = ;
2 3 4 12
1 1 1 1 19
c3 = + + + = ;
3 4 5 6 20
1 1 1 1 1 743
c4 = + + + + =
4 5 6 7 8 840
(b) We consider the left endpoint approximation to the integral of y = 1
over the interval [n, 2n]. Since the
$ 2n x
function y = 1
x is decreasing, the left endpoint approximation is greater than n dxx ; that is,
2n
dx 1 1 1 1
≤ ·1+ ·1+ · 1 + ··· + ·1
n x n n+1 n+2 2n − 1
 S E C T I O N 10.1 Sequences (LT SECTION 11.1) 25

y = 1x

1
1 1
2 3 1/n
x
0 1 2 3 n n+1

We express the right hand-side of the inequality in terms of cn , obtaining:

2n
dx 1
≤ cn −
n x 2n
$ 2n dx
We now consider the right endpoint approximation to the integral n x ; that is,
2n
1 1 1 dx
·1+ · 1 + ··· + ·1≤
n+1 n+2 2n n x
y

1
y=
x

1
n+1

x
0 n n+1

We express the left hand-side of the inequality in terms of cn , obtaining:

2n
1 dx
cn − ≤
n n x
Thus,
2n 2n
dx 1 dx 1
+ ≤ cn ≤ +
n x 2n n x n
(c) With
2n
dx 2n
= ln x|2n
n = ln 2n − ln n = ln = ln 2,
n x n
the result from part (b) becomes
1 1
ln 2 + ≤ cn ≤ ln 2 +
2n n
Because
1 1
lim = lim = 0
n→∞ 2n n→∞ n

it follows from the Squeeze Theorem that

lim cn = ln 2
n→∞

93. Let an = Hn − ln n, where Hn is the nth harmonic number:

1 1 1
Hn = 1 + + + ··· +
2 3 n
n+1
dx
(a) Show that an ≥ 0 for n ≥ 1. Hint: Show that Hn ≥ .
1 x
26 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

(b) Show that {an } is decreasing by interpreting an − an+1 as an area.

(c) Prove that lim an exists.
n→∞
This limit, denoted γ, is known as Euler’s Constant. It appears in many areas of mathematics, including
analysis and number theory, and has been calculated to more than 100 million decimal places, but it is still
not known whether γ is an irrational number. The first 10 digits are γ ≈ 0.5772156649.
solution
$ n+1
(a) Since the function y = 1x is decreasing, the left endpoint approximation to the integral 1 dxx is greater
than this integral; that is,
n+1
1 1 1 dx
1·1+ · 1 + · 1 + ··· + · 1 ≥
2 3 n 1 x
or
n+1
dx
Hn ≥
1 x
y

1
1
1
2
3 1/n
x
1 2 3 n n+1

Moreover, since the function y = 1

x is positive for x > 0, we have:
n+1 n
dx dx
≥
1 x 1 x
Thus,
n n
= ln x = ln n − ln 1 = ln n
dx
Hn ≥
1 x 1

and
an = Hn − ln n ≥ 0 for all n ≥ 1
(b) To show that {an } is decreasing, we consider the difference an − an+1 :
an − an+1 = Hn − ln n − (Hn+1 − ln(n + 1)) = Hn − Hn+1 + ln(n + 1) − ln n
 
1 1 1 1 1
= 1 + + ··· + − 1 + + ··· + + + ln(n + 1) − ln n
2 n 2 n n+1
1
=− + ln(n + 1) − ln n
n+1
$ n+1 $ n+1
Now, ln(n + 1) − ln n = n dxx , whereas n+1 1
is the right endpoint approximation to the integral n dx
x .
Recalling y = 1x is decreasing, it follows that
n+1
dx 1
≥
n x n+1
y

1
y=
x

1
n+1

x
n n+1
 S E C T I O N 10.2 Summing an Infinite Series (LT SECTION 11.2) 27

so

an − an+1 ≥ 0

(c) By parts (a) and (b), {an } is decreasing and 0 is a lower bound for this sequence. Hence 0 ≤ an ≤ a1 for
all n. A monotonic and bounded sequence is convergent, so limn→∞ an exists.

10.2 Summing an Infinite Series (LT Section 11.2)

Preliminary Questions
1. What role do partial sums play in defining the sum of an infinite series?
solution The sum of an infinite series is defined as the limit of the sequence of partial sums. If the limit of
this sequence does not exist, the series is said to diverge.
2. What is the sum of the following infinite series?
1 1 1 1 1
+ + + + + ···
4 8 16 32 64
solution This is a geometric series with c = 1
4 and r = 12 . The sum of the series is therefore
1 1
1
4
= 4
=
1− 1
2
1
2
2

3. What happens if you apply the formula for the sum of a geometric series to the following series? Is the
formula valid?

1 + 3 + 32 + 33 + 34 + · · ·

solution This is a geometric series with c = 1 and r = 3. Applying the formula for the sum of a geometric
series then gives

∞
1 1
3n = =−
n=0
1−3 2

Clearly, this is not valid: a series with all positive terms cannot have a negative sum. The formula is not valid
in this case because a geometric series with r = 3 diverges.
 ∞
1 1
4. Indicate whether or not the reasoning in the following statement is correct: 2
= 0 because 2 tends
n=1
n n
to zero.

solution This reasoning is not correct. Though the terms in the series do tend to zero, the general term in
the sequence of partial sums,
1 1 1
Sn = 1 + + + ··· + 2
22 32 n
is clearly larger than 1. The sum of the series therefore cannot be zero.
∞
1
5. Indicate whether or not the reasoning in the following statement is correct: √ converges because
n=1
n

1
lim √ = 0
n→∞ n

solution The reasoning is not correct. Although the general term of a convergent series must tend to zero,
∞
1
a series whose general term tends to zero need not converge. In the case of √ , the series diverges even
n=1
n
though its general term tends to zero.
28 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)


∞
6. Find an N such that S N > 25 for the series 2.
n=1

solution The Nth partial sum of the series is:


N
SN = 2 = 2
+ · · · + 2 = 2N
n=1 N


∞
7. Does there exist an N such that S N > 25 for the series 2−n ? Explain.

∞ n=1
1
solution The series 2−n is a convergent geometric series with the common ratio r = . The sum of the
n=1
2
series is:
1
S = 2
=1
1− 1
2

Notice that the sequence of partial sums {S N } is increasing and converges to 1; therefore S N ≤ 1 for all N.
Thus, there does not exist an N such that S N > 25.
8. Give an example of a divergent infinite series whose general term tends to zero.

∞
1 1
solution Consider the series 9
. The general term tends to zero, since lim 9
= 0. However, the Nth
n→∞ n 10
n=1 n
10

partial sum satisfies the following inequality:

1 1 1 N 9 1
SN = 9
+ 9
+ ··· + 9
≥ 9
= N 1− 10 = N 10
1 10 2 10 N 10 N 10

1 1
That is, S N ≥ N 10 for all N. Since lim N 10 = ∞, the sequence of partial sums S n diverges; hence, the series
N→∞
∞
1
9
diverges.
n=1 n
10

Exercises
1. Find a formula for the general term an (not the partial sum) of the infinite series.
1 1 1 1 1 5 25 125
(a) + + + + ··· (b) + + + + ···
3 9 27 81 1 2 4 8
1 22 33 44
(c) − + − + ···
1 2·1 3·2·1 4·3·2·1
2 1 2 1
(d) + + + + ···
12 + 1 22 + 1 32 + 1 42 + 1
solution
(a) The denominators of the terms are powers of 3, starting with the first power. Hence, the general term is:

1
an =
3n
(b) The numerators are powers of 5, and the denominators are the same powers of 2. The first term is a1 = 1
so,
 n−1
5
an =
2

(c) The general term of this series is

nn
an = (−1)n+1
n!
 S E C T I O N 10.2 Summing an Infinite Series (LT SECTION 11.2) 29

(d) Notice that the numerators of an equal 2 for odd values of n and 1 for even values of n. Thus,
⎧ 2
⎪
⎪
⎪
⎪
⎪
⎨ n2 + 1
odd n
an = ⎪⎪
⎪
⎪
⎪ 1
⎩ even n
n +1
2

The formula can also be rewritten as follows:

1 + (−1)2 +1
n+1

an =
n2 + 1
2. Write in summation notation:
1 1 1 1 1 1 1
(a) 1 + + + + ··· (b) + + + + ···
4 9 16 9 16 25 36
1 1 1
(c) 1 − + − + · · ·
3 5 7
125 625 3125 15,625
(d) + + + + ···
9 16 25 36
solution
1  1 ∞
(a) The general term is an = , n = 1, 2, 3, . . .; hence, the series is .
n2 n=1
n2
1 1
(b) The general term is an = 2 , n = 3, 4, 5, . . . or an = , n = 1, 2, 3, . . .; hence, the series is
n (n + 2)2
 1
∞ ∞
1
= .
n=3
n2
n=1 (n + 2)2
(−1)n+1  ∞
(−1)n+1
(c) The general term is an = , n = 1, 2, 3, . . .; hence, the series is .
2n − 1 n=1
2n − 1
5n 5n+2
(d) The general term is an = , n = 3, 4, 5, . . . or an = , n = 1, 2, 3, . . .; hence, the series is
n2 (n + 2)2

∞
5n 
∞
5n+2
= .
n=3
n2 n=1 (n + 2)2

In Exercises 3–6, compute the partial sums S 2 , S 4 , and S 6 .

1 1 1
3. 1 + 2
+ 2 + 2 + ···
2 3 4
solution
1 5
S2 = 1 + = ;
22 4
1 1 1 205
S4 = 1 + 2 + 2 + 2 = ;
2 3 4 144
1 1 1 1 1 5369
S6 = 1 + 2 + 2 + 2 + 2 + 2 =
2 3 4 5 6 3600

∞
4. (−1)k k−1
k=1

solution
1 1
S 2 = (−1)1 · 1−1 + (−1)2 · 2−1 = −1 + =− ;
2 2
1 1 1 1 1 7
S 4 = (−1)1 · 1−1 + (−1)2 · 2−1 + (−1)3 · 3−1 + (−1)4 · 4−1 = S 2 − + =− − + =− ;
3 4 2 3 4 12
7 7 1 1 37
S6 = − + (−1)5 · 5−1 + (−1)6 · 6−1 = − − + = −
12 12 5 6 60
30 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

1 1 1
5. + + + ···
1·2 2·3 3·4
solution
1 1 1 1 4 2
S2 = + = + = = ;
1·2 2·3 2 6 6 3
2 1 1 2 1 1 4
S 4 = S 2 + a3 + a4 = + + = + + = ;
3 3 · 4 4 · 5 3 12 20 5
4 1 1 4 1 1 6
S 6 = S 4 + a5 + a6 = + + = + + =
5 5 · 6 6 · 7 5 30 42 7
∞
1
6.
j=1
j!
solution
1 1 1 3
S2 = + =1+ = ;
1! 2! 2 2
1 1 3 1 1 41
S4 = S2 + + = + + = ;
3! 4! 2 6 24 24
1 1 41 1 1 1237
S6 = S4 + + = + + =
5! 6! 24 120 720 720
1 2 1 3
7. The series 1 + 5 + 5 + 5 + · · · converges to 4 . Calculate S N for N = 1, 2, . . . until you find an S N
1 5

that approximates 54 with an error less than 0.0001.

solution
S1 = 1
1 6
S2 = 1 + = = 1.2
5 5
1 1 31
S3 = 1 + + = = 1.24
5 25 25
1 1 1 156
S3 = 1 + + + = = 1.248
5 25 125 125
1 1 1 1 781
S4 = 1 + + + + = = 1.2496
5 25 125 625 625
1 1 1 1 1 3906
S5 = 1 + + + + + = = 1.24992
5 25 125 625 3125 3125
Note that
1.25 − S 5 = 1.25 − 1.24992 = 0.00008 < 0.0001
1 1 1 1
8. The series − + − + · · · is known to converge to e−1 (recall that 0! = 1). Calculate S N for
0! 1! 2! 3!
N = 1, 2, . . . until you find an S N that approximates e−1 with an error less than 0.001.
solution The general term of the series is
(−1)n−1
an =
(n − 1) !
thus, the Nth partial sum of the series is

N N
(−1)n−1 1 1 1 (−1)N−1
SN = an = = − + − ··· +
n=1 n=1
(n − 1) ! 0! 1! 2! (N − 1) !

Using a calculator we find e−1 = 0.367879. Working sequentially, we find

1
S1 = =1
0!
1
S 2 = S 1 + a2 = 1 − =0
1!
 S E C T I O N 10.2 Summing an Infinite Series (LT SECTION 11.2) 31

1 1
S 3 = S 2 + a3 = 0 + = = 0.5
2! 2
1
S 4 = S 3 + a4 = 0.5 − = 0.333333
3!
1
S 5 = S 4 + a5 = 0.333333 + = 0.375
4!
1
S 6 = S 5 + a6 = 0.375 − = 0.366667
5!
1
S 7 = S 6 + a7 = 0.366667 + = 0.368056
6!
Note that
|S 7 − e−1 | = 1.76 × 10−4 < 10−3

In Exercises 9 and 10, use a computer algebra system to compute S 10 , S 100 , S 500 , and S 1000 for the series.
Do these values suggest convergence to the given value?
9.
π−3 1 1 1 1
= − + − + ···
4 2 · 3 · 4 4 · 5 · 6 6 · 7 · 8 8 · 9 · 10
solution Write
(−1)n+1
an =
2n · (2n + 1) · (2n + 2)
Then

N
SN = an
i=1

Computing, we find
π−3
≈ 0.0353981635
4
S 10 ≈ 0.03535167962
S 100 ≈ 0.03539810274
S 500 ≈ 0.03539816290
S 1000 ≈ 0.03539816334
π−3
It appears that S N → 4 .
10.
π4 1 1 1
= 1 + 4 + 4 + 4 + ···
90 2 3 4
solution Write
N
1
SN = 4
i=1
i
Computing, we find
π4
≈ 1.082323234
90
S (10) ≈ 1.082036583
S (100) ≈ 1.082322905
S (500) ≈ 1.082323231
S (1000) ≈ 1.082323233
π4
It appears that S N → 90 .
32 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

11. Calculate S 3 , S 4 , and S 5 and then find the sum of the telescoping series
∞  
1 1
−
n=1
n+1 n+2

solution
     
1 1 1 1 1 1 1 1 3
S3 = − + − + − = − = ;
2 3 3 4 4 5 2 5 10
 
1 1 1 1 1
S4 = S3 + − = − = ;
5 6 2 6 3
 
1 1 1 1 5
S5 = S4 + − = − =
6 7 2 7 14
The general term in the sequence of partial sums is
       
1 1 1 1 1 1 1 1 1 1
SN = − + − + − + ··· + − = −
2 3 3 4 4 5 N+1 N+2 2 N+2
thus,
 
1 1 1
S = lim S N = lim − =
N→∞ N→∞ 2 N+2 2

The sum of the telescoping series is therefore 12 .


∞
1
12. Write as a telescoping series and find its sum.
n=3
n(n − 1)

solution By partial fraction decomposition

1 1 1
= −
n(n − 1) n − 1 n
so

∞ ∞ 
 
1 1 1
= −
n=3
n(n − 1) n=3 n − 1 n

The general term in the sequence of partial sums for this series is
       
1 1 1 1 1 1 1 1 1 1
SN = − + − + − + ··· + − = −
2 3 3 4 4 5 N−1 N 2 N
thus,
 
1 1 1
S = lim S N = lim − =
N→∞ N→∞ 2 N 2

∞
1
13. Calculate S 3 , S 4 , and S 5 and then find the sum 2−1
using the identity
4n
n=1 
1 1 1 1
= −
4n2 − 1 2 2n − 1 2n + 1

solution
       
1 1 1 1 1 1 1 1 1 1 1 3
S3 = − + − + − = 1− = ;
2 1 3 2 3 5 2 5 7 2 7 7
   
1 1 1 1 1 4
S4 = S3 + − = 1− = ;
2 7 9 2 9 9
   
1 1 1 1 1 5
S5 = S4 + − = 1− =
2 9 11 2 11 11
 S E C T I O N 10.2 Summing an Infinite Series (LT SECTION 11.2) 33

The general term in the sequence of partial sums is

         
1 1 1 1 1 1 1 1 1 1 1 1 1 1
SN = − + − + − + ··· + − = 1−
2 1 3 2 3 5 2 5 7 2 2N − 1 2N + 1 2 2N + 1
thus,
 
1 1 1
S = lim S N = lim 1− =
N→∞ N→∞ 2 2N + 1 2

∞
1
14. Use partial fractions to rewrite as a telescoping series and find its sum.
n=1
n(n + 3)
solution By partial fraction decomposition
1 A B
= +
n (n + 3) n n+3
clearing denominators gives
1 = A (n + 3) + Bn
Setting n = 0 yields A = 13 , while setting n = −3 yields B = − 13 . Thus,
 
1 1 1 1
= −
n(n + 3) 3 n n + 3
and

∞ 
∞  
1 1 1 1
= −
n=1
n(n + 3) n=1 3 n n + 3

The general term in the sequence of partial sums for the series on the right-hand side is
           
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
SN = 1− + − + − + − + − + −
3 4 3 2 5 3 3 6 3 4 7 3 5 8 3 6 9
     
1 1 1 1 1 1 1 1 1
+··· + − + − + −
3 N−2 N+1 3 N−1 N+2 3 N N+3
     
1 1 1 1 1 1 1 11 1 1 1 1
= 1+ + − + + = − + +
3 2 3 3 N+1 N+2 N+3 18 3 N + 1 N + 2 N + 3
Thus,
%  &
11 1 1 1 1 11
lim S N = lim − + + =
N→∞ N→∞ 18 3 N + 1 N + 2 N + 3 18
and

∞
1 11
=
n=1
n(n + 3) 18

1 1 1
15. Find the sum of + + + ··· .
1·3 3·5 5·7
solution We may write this sum as

∞ 
∞  
1 1 1 1
= −
n=1
(2n − 1)(2n + 1) n=1 2 2n − 1 2n + 1

The general term in the sequence of partial sums is

         
1 1 1 1 1 1 1 1 1 1 1 1 1 1
SN = − + − + − + ··· + − = 1−
2 1 3 2 3 5 2 5 7 2 2N − 1 2N + 1 2 2N + 1
thus,
 
1 1 1
lim S N = lim 1− =
N→∞ N→∞ 2 2N + 1 2
34 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

and

∞
1 1
=
n=1
(2n − 1)(2n + 1) 2


∞
16. Find a formula for the partial sum S N of (−1)n−1 and show that the series diverges.
n=1

solution The partial sums of the series are:

S 1 = (−1)1−1 = 1;
S 2 = (−1)0 + (−1)1 = 1 − 1 = 0;
S 3 = (−1)0 + (−1)1 + (−1)2 = 1;
S 4 = (−1)0 + (−1)1 + (−1)2 + (−1)3 = 0 · · ·

In general,
'
1 if N odd
SN =
0 if N even
Because the values of S N alternate between 0 and 1, the sequence of partial sums diverges; this, in turn,

∞
implies that the series (−1)n−1 diverges.
n=1

In Exercises 17–22, use the nth Term Divergence Test (Theorem 4) to prove that the following series diverge.

∞
n
17.
n=1
10n + 12
n
solution The general term, , has limit
10n + 12
n 1 1
lim = lim =
n→∞ 10n + 12 n→∞ 10 + (12/n) 10
Since the general term does not tend to zero, the series diverges.
∞
n
18. √
n=1 n +1
2

n
solution The general term, √ , has limit
n2 + 1

n n2 1
lim √ = lim = lim =1
n→∞ n2 +1 n→∞ n2 + 1 n→∞ 1 + (1/n2 )

Since the general term does not tend to zero, the series diverges.
0 1 2 3
19. − + − + · · ·
1 2 3 4
solution The general term an = (−1)n−1 n−1 n does not tend to zero. In fact, because limn→∞
n−1
n = 1,
limn→∞ an does not exist. By Theorem 4, we conclude that the given series diverges.
∞
20. (−1)n n2
n=1

solution The general term an = (−1)n n2 does not tend to zero. In fact, because limn→∞ n2 = ∞, limn→∞ an
does not exist. By Theorem 3, we conclude that the given series diverges.
1 1 1
21. cos + cos + cos + · · ·
2 3 4
solution The general term an = cos n+1 1
tends to 1, not zero. By Theorem 4, we conclude that the given
series diverges.
 S E C T I O N 10.2 Summing an Infinite Series (LT SECTION 11.2) 35


∞ √
22. 4n2 + 1 − n
n=0
solution The general term of the series satisfies
√ √
4n2 + 1 − n > 4n2 − n = n

Thus the general term tends to infinity. The series diverges by Theorem 2.
In Exercises 23–38, either use the formula for the sum of a geometric series to find the sum, or state that the
series diverges.
1 1 1
23. 1 + + + + ···
6 36 216
solution This is a geometric series with
1
c = 1 and r=
6
so its sum is
c 1 1 6
= = =
1−r 1− 1
6
5
6
5

43 44 45
24. + + + ···
53 54 55
solution This is a geometric series with
43 4
c= and r=
53 5
so its sum is
c 43 /53 43 64
= = 3 =
1−r 1− 5 4 5 − 4 · 52 25

7 7 7 7
25. + + + + ···
3 32 33 34
solution This is a geometric series with
7 1
c= and r=
3 3
so its sum is
c 7/3 7 7
= = =
1 − r 1 − 13 3−1 2
 2  3  4
7 7 7 7
26. + + + + ···
3 3 3 3
solution This is a geometric series with
7 7
c= and r=
3 3
Since r > 1, the series diverges.
∞  −n
3
27.
n=3
11
solution Rewrite this series as
∞ 
 n
11
n=3
3

11
This is a geometric series with r = > 1, so it is divergent.
3
36 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)


∞
7 · (−3)n
28.
n=2
5n
3
solution This is a geometric series with c = 7 and r = − , starting at n = 2. Its sum is thus
5
cr2 7 · (9/25) 63 5 63
= = · =
1−r 1 − 35 25 8 40

∞  n
4
29. −
n=−4
9
4
solution This is a geometric series with c = 1 and r = − , starting at n = −4. Its sum is thus
9
cr−4 c 1 95 59,049
= = = =
1 − r r4 − r5 44
+ 45 9·4 +4
4 5 3328
94 95

∞  n
 π
30.
n=0
e
solution Since π > e, this is a geometric series with r > 1, so it diverges.
∞
31. e−n
n=1
solution Rewrite the series as
∞  n
 1
n=1
e

to recognize it as a geometric series with c = 1

e and r = 1e . Thus,

∞ 1
1
e−n = e
=
n=1 1− 1
e
e−1


∞
32. e3−2n
n=2
solution Rewrite the series as

∞ 
∞  n
3 −2n 1
e e = e3

n=2 n=2
e2

1 2
to recognize it as a geometric series with c = e3 e2
= 1
e and r = 1
e2
. Thus,

∞ 1
e
e3−2n = e
=
n=2 1− 1
e2
e2 −1


∞
8 + 2n
33.
n=0
5n
solution Rewrite the series as
  n  ∞  n
8  2n 
∞ ∞ ∞
1 2
n
+ n
= 8· +
n=0
5 n=0
5 n=0
5 n=0
5

1 0
which is a sum of two geometric series. The first series has c = 8 5 = 8 and r = 1
5; the second has
 0
c = 25 = 1 and r = 25 . Thus,
 ∞  n
1 8 8
8· = = 4 = 10
n=0
5 1 − 1
5 5
 S E C T I O N 10.2 Summing an Infinite Series (LT SECTION 11.2) 37

∞  n
 2 1 1 5
= = =
n=0
5 1− 2
5
3
5
3

and

∞
8 + 2n 5 35
= 10 + =
n=0
5n 3 3


∞
3(−2)n − 5n
34.
n=0
8n
solution Rewrite the series as

∞
3(−2)n − 5n 
∞
3(−2)n 
∞
5n
= −
n=0
8n n=0
8n n=0
8n

which is a difference of two geometric series. The first has c = 3 and r = − 14 ; the second has c = 1 and r = 58 .
Thus
∞
3(−2)n 3 12
= =
n=0
8n
1+ 41 5

∞
5n 1 8
= =
n=0
8n 1− 5
8
3

so that

∞
3(−2)n − 5n 12 8 4
= − =−
n=0
8n 5 3 15

5 5 5
35. 5 − + − + ···
4 42 43
solution This is a geometric series with c = 5 and r = − 41 . Thus,
∞  n
1 5 5 5
5· − =  = = 5 =4
n=0
4 1 − −4 1 1 + 1
4 4

23 24 25 26
36. + 2 + 3 + 4 + ···
7 7 7 7
solution This is a geometric series with c = and r = 27 . Thus,
8
7

∞  n 8 8
8 2 8
· = 7 2 = 75 =
n=0
7 7 1− 7 7
5

7 49 343 2401
37. − + − + ···
8 64 512 4096
solution This is a geometric series with c = 78 and r = − 78 . Thus,
 ∞  n 7 7
7 7 7
· − = 8 8
= 15 =
n=0
8 8 1− − 7
8
15
8

25 5 3 9 27
38. + +1+ + + + ···
9 3 5 25 125
solution This appears to be a geometric series with
25 3
c= and r=
9 5
so its sum is
c 25/9 25 5 125
= = · =
1−r 1− 53 9 2 18
38 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

In Exercises 39–44, determine a reduced fraction that has this decimal expansion.
39. 0.222 . . .
solution The decimal may be regarded as a geometric series:

2 2 2  2 ∞
0.222 . . . = + + ... =
10 100 1000 n=1
10n

The series has first term 2

10 = 1
5 and ratio 1
10 , so its sum is
1/5 1 10 2
0.222 . . . = = · =
1 − 1/10 5 9 9

40. 0.454545 . . .
solution The decimal may be regarded as a geometric series:

45 45 45  45 ∞
0.454545 . . . = + + ... =
100 10000 1000000 n=1
102n

The series has first term 45

100 = 9
20 and ratio 1
100 , so its sum is
9/20 9 100 5
0.454545 . . . = = · =
1 − 1/100 20 99 11

41. 0.313131 . . .
solution The decimal may be regarded as a geometric series:

31 31 31  31 ∞
0.313131 . . . = + + ... =
100 10000 1000000 n=1
102n
31 1
The series has first term 100 and ratio 100 , so its sum is
31/100 31 100 31
0.313131 . . . = = · =
1 − 1/100 100 99 99

42. 0.217217217 . . .
solution The decimal may be regarded as a geometric series:

217 217 217  217

∞
0.217217217 . . . = 3
+ 6 + 9 ... =
10 10 10 n=1
103n
217 1
The series has first term 103
and ratio 103
, so its sum is

217/103 217 103 217

0.217217217 . . . = = 3 · =
1 − 1/103 10 999 999

43. 0.123333333 . . .
solution The decimal may be regarded as a constant plus a geometric series:
0.123333333 . . . = 0.12 + 0.003 + 0.0003 + 0.00003 + · · ·
12 3 3 3
= + + + + ···
100 103 104 105
12  3
∞
= +
100 n=1 10n+2
3 1
The series has first term 1000 and ratio 10 , so the sum is
12 3/1000 12 3 111
+ = + =
100 1 − 10
1 100 900 900
 S E C T I O N 10.2 Summing an Infinite Series (LT SECTION 11.2) 39

44. 0.808888888 . . .
solution The decimal may be regarded as a constant plus a geometric series:
0.808888888 . . . = 0.80 + 0.008 + 0.0008 + 0.00008 + · · ·
80 8 8 8
= + + + + ···
100 103 104 105
80  8
∞
= +
100 n=1 10n+2
8 1
The series has first term 1000 and ratio 10 , so the sum is
80 8/1000 80 8 728
+ = + =
100 1 − 10
1 100 900 900

45. Verify that 0.999999 · · · = 1 by expressing the left side as a geometric series and determining the sum of
the series.
solution The decimal may be regarded as a geometric series:

9 9 9  9 ∞
0.999999 · · · = + + + ··· =
10 100 1000 n=1
10n
9 1
This series has first term 10 and ratio 10 , so its sum is
9/10 9/10
0.999999 · · · = = =1
1 − 10
1 9/10

46. The repeating decimal

0.012345678901234567890123456789 . . .
can be expressed as a fraction with denominator 1,111,111,111. What is the numerator?
solution The decimal may be regarded as a geometric series:
123456789 123456789 123456789
0.012345678901234567890123456789 · · · = + + + ···
1010 1020 1030
123456789 1
This series has first term 1010
and ratio 1010
, so its sum is

123456789/1010 123456789 13,717,421

= =
1 − 1010
1 10 − 1
10 1,111,111,111

The numerator is 13,717,421.

47. Which of the following are not geometric series?
∞
7n ∞
1
(a) n
(b) 4
n=0
29 n=3
n
∞
n2 ∞
(c) n
(d) π−n
n=0
2 n=5

solution
∞ ∞  n
7n 7 7
(a) n = : this is a geometric series with common ratio r = .
n=0
29 n=0
29 29
(b) The ratio between two successive terms is

an+1
1
n4  n 4
(n+1)4
= = =
an 1
n4
(n + 1)4 n+1
∞
1
This ratio is not constant since it depends on n. Hence, the series 4
is not a geometric series.
n=3
n
40 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

(c) The ratio between two successive terms is

(n+1)2  2
an+1 2n+1 (n + 1)2 2n 1 1
= = · = 1 + ·
an n2 n2 2n+1 n 2
2n


∞ 2
n
This ratio is not constant since it depends on n. Hence, the series is not a geometric series.
n=0
2n
∞ ∞  n
1 1
(d) π−n = : this is a geometric series with common ratio r = .
n=5 n=5
π π

∞
1
48. Use the method of Example 10 to show that 1/3
diverges.
k=1
k
1
solution Each term in the Nth partial sum is greater than or equal to 1
, hence:
N3
1 1 1 1 1 1 1 1 1
SN = + + + ··· + ≥ 1/3 + 1/3 + 1/3 + · · · + 1/3 = N · 1/3 = N 2/3
11/3 21/3 31/3 N 1/3 N N N N N
Since lim N 2/3 = ∞, it follows that
N→∞

lim S N = ∞
N→∞

∞
1
Thus, the series 1/3
diverges.
k=1
k

∞ 
∞
49. Prove that if an converges and bn diverges, then
n=1 n=1

∞
(an + bn ) diverges. Hint: If not, derive a contradiction by writing
n=1


∞ 
∞ 
∞
bn = (an + bn ) − an
n=1 n=1 n=1
# #∞ #∞
solution Suppose to the contrary that ∞ n=1 an converges, n=1 bn diverges, but n=1 (an + bn ) converges.
Then by the Linearity of Infinite Series, we have

∞ 
∞ 
∞
bn = (an + bn ) − an
n=1 n=1 n=1
#∞
so that n=1 bn converges, a contradiction.

∞
9n + 2n
50. Prove the divergence of .
n=0
5n

solution Note that this is the sum of two infinite series:


∞
9n + 2n 
∞
9n 
∞
2n
= +
n=0
5n n=0
5n n=0
5n

The first of these is a geometric series with r = 95 > 1, so diverges, while the second is a geometric series
with r = 25 < 1, so converges. By the previous exercise, the sum of the two also diverges.

51. Give a counterexample to show that each of the following statements is false.

∞
(a) If the general term an tends to zero, then an = 0.
n=1
(b) The Nth partial sum of the infinite series defined by {an } is aN .
 S E C T I O N 10.2 Summing an Infinite Series (LT SECTION 11.2) 41


∞
(c) If an tends to zero, then an converges.
n=1

∞
(d) If an tends to L, then an = L.
n=1

solution
(a) If the general term an tends to zero, the series may or may not converge. Even if the series converges, it
may not converge to zero. For example, with the harmonic series, where an = 1n , we have lim an = 0, but
n→∞
∞
an diverges.
n=1

∞
(b) The Nth partial sum of the series an is S N = a1 + a2 + · · · + aN rather than aN . For example, take the
n=1
infinite series defined by an = 1 for all n. Then aN = 1, but the Nth partial sum is N.
(c) If the general term an tends to zero, the series may or may not converge. See part (a) for an example.
(d) If L  0, then the series diverges by the nth Term Divergence Test. If L = 0, the series may or may not
converge, and even if it does converge, it may not converge to L = 0. For an example of the first case, take
the series defined by an = 1 for all n; for the second, take the harmonic series.

∞
2
52. Suppose that an is an infinite series with partial sum S N = 5 − .
n=1
N2

10 
16
(a) What are the values of an and an ?
n=1 n=5

(b) What is the value of a3 ?

(c) Find a general formula for an .
∞
(d) Find the sum an .
n=1

solution
(a)


10
2 249
an = S 10 = 5 − 2
= ;
n=1 10 50


16    
2 2 2 2 494
an = 1 + · · · + a16 − 1 + a2 + a3 = S 16 − S 3 = 5 − 2 − 5 − 2 = −
(a ) (a ) =
n=4 16 3 9 256 2304

(b)
  
2 2 1 2 5
a3 = (a1 + a2 + a3 ) − (a1 + a2 ) = S 3 − S 2 = 5 − 2 − 5 − 2 = − =
3 2 2 9 18

(c) Since an = S n − S n−1 , we have:

   
2 2 2 2
an = S n − S n−1 = 5 − 2 − 5 − = −
n (n − 1)2 (n − 1)2 n2
 
2 n2 − (n − 1)2 2 n2 − n2 + 2n − 1 2 (2n − 1)
= = =
(n (n − 1)) 2
(n (n − 1)) 2
n2 (n − 1)2

∞
(d) The sum an is the limit of the sequence of partial sums {S N }. Hence:
n=1


∞  
2
an = lim S N = lim 5 − 2 = 5
n=1
N→∞ N→∞ N
42 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

53. Consider the archery competition in Example 6.

(a) Assume that Nina goes first. Let pn represent the probability that Brook wins on his nth turn. Give an
expression for pn .
(b) Use the result from (a) and a geometric series to determine the probability that Brook wins when Nina
goes first.
(c) Now assume that Brook goes first. Use a geometric series to compute the probability that Brook wins
the competition.
solution
(a) The probability pn that Brook wins on his nth turn when Nina goes first is the probability that Nina misses
on her first shot, then both players miss on n − 1 consecutive shots, and then Brook hits. This probability is

pn = 0.55 · (0.48 · 0.55)n−1 · 0.52 = 0.286 · 0.264n−1

(b) pn is a geometric series with first term (n = 1) of 0.286 and a ratio of 0.264, so its sum is

∞
0.286 0.286
p= pn = = ≈ 0.389
n=1
1 − 0.264 0.736

(c) Let qn be the probability that Brook wins on his nth turn when he goes first. As before, this is the
probability that both players miss on their first n − 1 turns times the probability that Brook hits on his nth
turn. The probability that both players miss on any given turn is (0.48)(0.55), and the probability that Brook
hits on his nth turn is 0.52. Thus

qn = (0.48 · 0.55)n−1 · 0.52 = 0.52 · 0.264n−1

qn is a geometric series with first term (n = 1) of 0.52 and a ratio of 0.264, so its sum is

∞
0.52 0.52
q= qn = = ≈ 0.707
n=1
1 − 0.264 0.736

54. Consider the archery competition in Example 6. Assume that Nina’s probability of hitting the bull’s-eye
on a turn is 0.45 and that Brook’s probability is p. Assume that Nina goes first. For what value of p do both
players have a probability of 1/2 of winning the competition?
solution If Nina wins with probability 1/2, clearly so does Brook. The probability of Nina winning on
her nth turn is, as in the previous exercise, the probability that both players miss on all prior turns times
the probability that she hits on her nth turn. The probability that both players miss on any given turn is
0.55(1 − p), and the probability that Nina hits on her nth turn is 0.45. Thus Nina’s win probability on the nth
turn is

(0.55(1 − p))n−1 · 0.45.

Therefore her total win probability is a geometric series with first term (n = 1) of 0.45 and a ratio of
0.55(1 − p), so its sum is

∞
0.45
0.45(0.55(1 − p))n−1 =
n=1
1 − 0.55(1 − p)

Solving 0.45
1−0.55(1−p) = 0.5 gives p ≈ 0.8182.
55. Compute the total area of the (infinitely many) triangles in Figure 4.
y
1
2

x
1 1 1 1 1
16 8 4 2
FIGURE 4
 S E C T I O N 10.2 Summing an Infinite Series (LT SECTION 11.2) 43

solution The area of a triangle with base B and height H is A = 12 BH. Because all of the triangles in
Figure 4 have height 12 , the area of each triangle equals one-quarter of the base. Now, for n ≥ 0, the nth
triangle has a base which extends from x = 2n+1
1
to x = 21n . Thus,

1 1 1 1 1
B= n
− n+1 = n+1 and A= B = n+3
2 2 2 4 2
The total area of the triangles is then given by the geometric series

∞ ∞  n 1
1 1 1 8 1
= = =
n=0
2n+3
n=0
8 2 1 − 1
2
4

56. The winner of a lottery receives m dollars at the end of each year for N years. The present value (PV) of

N
this prize in today’s dollars is PV = m(1 + r)−i , where r is the interest rate. Calculate PV if m = $50,000,
i=1
r = 0.06 (corresponding to 6%), and N = 20. What is PV if N = ∞?
solution For the given values r, m and N, we have

 i 
  50 21
20 20
50 1−
−i 53
PV = 50,000(1 + 0.06) = 50,000 = 50,000 = $623,496.06
i=1 i=1
53 1 − 50
53

If we extend the payments forever, then N = ∞ and

 i 

∞ 
∞
50 50,000 50
−i 53
PV = 50,, 000(1 + 0.06) = 50,000 = = $833,333.33
i=1 i=1
53 1− 50
53

57. If a patient takes a dose of D units of a particular drug, the amount of the dosage that remains in the
patient’s bloodstream after t days is De−kt , where k is a positive constant depending on the particular drug.
(a) Show that if the patient takes a dose D every day for an extended period, the amount of drug in the
De−k
bloodstream approaches R = .
1 − e−k
(b) Show that if the patient takes a dose D once every t days for an extended period, the amount of drug in
De−kt
the bloodstream approaches R = .
1 − e−kt
(c) Suppose that it is considered dangerous to have more than S units of the drug in the bloodstream. What
is the minimal time between doses that is safe? Hint: D + R ≤ S .
solution
t 
t
(a) The total amount remaining after t days is De−kn = D e−kn , so that as t → ∞, the amount
n=1
n=1

∞  n  n−1
−kn −kn −k
approaches D e . Since e = D e = D e−k −k
· e , this is a geometric series with initial term
n=1
De−k and ratio e , so the sum of the series is
−k

De−k
R=
1 − e−k

∞
(b) In the equation for part (a), R = D e−kn , replace n by nt; this gives the series where dosages are given
n=1
only on days that are a multiple of t; that is, only every t days. Then

∞ 
∞ 
n De−kt
R=D e−knt = D e−kt =
n=1 n=1
1 − e−kt

summing the geometric series as in part (a).

44 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

(c) Since we want the total amount in the bloodstream to be at most S , and since the amount in the blood-
stream is at a maximum immediately after a dose is given, we want D + R ≤ S . From part (b), we get

De−kt
D+R= D+ ≤S
1 − e−kt
Multiply both sides of this inequality by the positive number 1 − e−kt , giving

D(1 − e−kt ) + De−kt ≤ S (1 − e−kt ), or D ≤ S (1 − e−kt )

Simplifying gives e−kt ≤ S −D

S . Take logs and divide both sides by −k (remembering to reverse the inequality
on dividing by a negative number), giving
1 S −D 1 S
t ≥ − log = log
k S k S −D
So the minimum amount of time between doses is the expression on the right-hand side of the final inequality.
58. In economics, the multiplier effect refers to the fact that when there is an injection of money to con-
sumers, the consumers spend a certain percentage of it. That amount recirculates through the economy and
adds additional income, which comes back to the consumers and of which they spend the same percentage.
This process repeats indefinitely, circulating additional money through the economy. Suppose that in order to
stimulate the economy, the government institutes a tax cut of $10 billion. If taxpayers are known to save 10%
of any additional money they receive, and to spend 90%, how much total money will be circulated through
the economy by that single $10 billion tax cut?
solution Of the initial $10 billion, 90% of it, or 0.9 × $10 = $9 billion, will be spent. Of that amount, an
additional 90% of it, or 0.9 × $9 = $8.1 billion, will be spent; this process continues. At each step, starting
with step t = 0 initially, the amount spent is 10 × 0.9t , so the total amount spent is the sum of a geometric
series with initial value 10 and ratio 0.9:

∞
10 10
T= 10 × 0.9t = = = 100
t=0
1 − 0.9 0.1

The initial $10 billion tax cut will result in $100 billion of additional consumer spending.
59. Find the total length of the infinite zigzag path in Figure 5 (each zag occurs at an angle of π4 ).

/4 /4
1
FIGURE 5

solution Because the angle at the lower left in Figure 5 has measure π4 and each zag in the path occurs at
an angle of π4 , every triangle in the figure is an isosceles right triangle. Accordingly, the length of each new
segment in the path is √12 times the length of the previous segment. Since the first segment has length 1, the
total length of the path is
∞  n √
1 1 2 √
√ = = √ =2+ 2
n=0 2 1 − √2
1
2−1

∞
1
60. Evaluate . Hint: Find constants A, B, and C such that
n=1
n(n + 1)(n + 2)
1 A B C
= + +
n(n + 1)(n + 2) n n + 1 n + 2
∞
1
and use the result to evaluate .
n=1
n(n + 1)(n + 2)
 S E C T I O N 10.2 Summing an Infinite Series (LT SECTION 11.2) 45

solution By partial fraction decomposition

1 A B C
= + +
n(n + 1)(n + 2) n n + 1 n + 2
clearing denominators then gives
1 = A (n + 1) (n + 2) + Bn (n + 2) + Cn (n + 1)
Setting n = 0 now yields A = 12 , while setting n = −1 yields B = −1 and setting n = −2 yields C = 12 . Thus,
1 1  
1 1 1 1 2 1
= −2
+ 2
= − +
n (n + 1) (n + 2) n n + 1 n + 2 2 n n + 1 n + 2
and

∞ 
∞  
1 1 1 2 1
= − +
n=1
n(n + 1)(n + 2) n=1 2 n n + 1 n + 2

The general term of the sequence of partial sums for the series on the right-hand side is
         
1 2 1 1 1 2 1 1 1 2 1 1 1 2 1 1 1 2 1
SN = 1− + + − + + − + + + + + − +
2 2 3 2 2 3 4 2 3 4 5 2 4 5 6 2 5 6 7
     
1 1 2 1 1 1 2 1 1 1 2 1
+··· + − + + − + + − +
2 N−2 N−1 N 2 N−1 N N+1 2 N N+1 N+2
 
1 1 1 1
= − +
2 2 N+1 N+2
Thus,

∞  
1 1 1 1 1 1
= lim S N = lim − + =
n=1
n(n + 1)(n + 2) N→∞ N→∞ 2 2 N+1 N+2 4

61. Show that if a is a positive integer, then


∞  
1 1 1 1
= 1 + + ··· +
n=1
n(n + a) a 2 a

solution By partial fraction decomposition

1 A B
= +
n (n + a) n n+a
clearing the denominators gives
1 = A(n + a) + Bn
Setting n = 0 then yields A = 1a , while setting n = −a yields B = − 1a . Thus,
1 1  
1 1 1 1
= −
a a
= −
n (n + a) n n + a a n n + a
and

∞ 
∞  
1 1 1 1
= −
n=1
n(n + a) n=1 a n n + a

For N > a, the Nth partial sum is

   
1 1 1 1 1 1 1 1 1
SN = 1 + + + ··· + − + + + ··· +
a 2 3 a a N+1 N+2 N+3 N+a
Thus,

∞  
1 1 1 1 1
= lim S N = 1 + + + ··· +
n=1
n(n + a) N→∞ a 2 3 a
46 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

62. A ball dropped from a height of 10 ft begins to bounce vertically. Each time it strikes the ground, it
returns to two-thirds of its previous height. What is the total vertical distance traveled by the ball if it bounces
infinitely many times?
solution The distance traveled by the ball is shown in the accompanying figure:
h = 10

2 2
h h
3 3

( 23 ) h ( 23 ) h
2 2

The total distance d traveled by the ball is given by the following infinite sum:
 2  3 ⎛  2  3 ⎞ ∞  n
2 2 2 ⎜⎜⎜ 2 2 2 ⎟⎟⎟  2
d =h+2· h+2· h+2· h + · · · = h + 2h ⎜⎜⎝ + + + · · · ⎟⎟⎠ = h + 2h
3 3 3 3 3 3 n=1
3

We use the formula for the sum of a geometric series to compute the sum of the resulting series:
 1
2
3
d = h + 2h · = h + 2h(2) = 5h
1− 2
3

With h = 10 feet, it follows that the total distance traveled by the ball is 50 feet.
63. In this exercise, we resolve the paradox of Gabriel’s Horn (Example 3 in Section 7.7 and Example
7 in Section 8.2). Recall that the horn is the surface formed by rotating y = 1x for x ≥ 1 around the x-axis.
The surface encloses a finite volume and has an infinite surface area. Thus, apparently we can fill the surface
with a finite volume of paint, but an infinite volume of paint is required to paint the surface.
(a) Explain that if we can fill the horn with paint, then the paint must be Magic Paint that can be spread
arbitrarily thin, thinner than the thickness of the molecules in normal paint.
(b) Explain that if we use Magic Paint, then we can paint the surface of the horn with a finite volume of
paint, in fact with just a milliliter of it. Hint: A geometric series helps here. Use half of a milliliter to paint
that part of the surface between x = 1 and x = 2.
solution
(a) If we can fill the horn with paint, then clearly the surface will be painted as well. Since the surface area
is infinite but the amount of paint used is finite, the thickness of the portion of the paint on the surface must
be arbitrarily small. Thus it must be Magic Paint.
(b) Since we can spread the paint arbitrarily thinly, start by using half a milliliter of paint to paint the part of
the surface between x = 1 and x = 2. For the part between x = 2 and x = 3, spread the paint so that we need
only one-quarter of a milliliter. Between x = 3 and x = 4, use only one-eighth of a milliliter, and so forth.
Then the total amount of Magic Paint used to paint the surface is
1 1 1 1/2
+ + + ··· = =1
2 22 23 1 − 12
So using Magic Paint, we can in fact paint the surface with only 1 milliliter of paint.
64. A unit square is cut into nine equal regions as in Figure 6(A). The central subsquare is painted red. Each
of the unpainted squares is then cut into nine equal subsquares and the central square of each is painted red
as in Figure 6(B). This procedure is repeated for each of the resulting unpainted squares. After continuing
this process an infinite number of times, what fraction of the total area of the original square is painted?

(A) (B)
FIGURE 6
 S E C T I O N 10.2 Summing an Infinite Series (LT SECTION 11.2) 47

solution Initially, 19 of the square is painted. Of the remaining unpainted 89 , another 19 is painted, leaving
 2
81 = 9
64 8
unpainted. Another 19 of that is painted in the third step. So eventually, the amount painted is


∞  n
1 8 1/9 1/9
= = =1
n=0
9 9 1 − 8/9 1/9

In the limit, the entire square is painted.

An alternative approach to this problem is simply to consider the amount of unpainted area remaining
 2  n  n
after each step. After the first step, it is 89 ; after the second, 89 ; after the nth it is 89 . Since lim 89 = 0,
n→∞
none of the square remains unpainted.
∞
65. Let {bn } be a sequence and let an = bn − bn−1 . Show that an converges if and only if lim bn exists.
n→∞
n=1


∞
solution Let an = bn − bn−1 . The general term in the sequence of partial sums for the series an is then
n=1

S N = (b1 − b0 ) + (b2 − b1 ) + (b3 − b2 ) + · · · + (bN − bN−1 ) = bN − b0


∞ 
∞
Now, if lim bN exists, then so does lim S N and an converges. On the other hand, if an converges,
N→∞ N→∞
n=1 n=1

∞
then lim S N exists, which implies that lim bN also exists. Thus, an converges if and only if lim bn exists.
N→∞ N→∞ n→∞
n=1

66. Assumptions Matter Show, by giving counterexamples, that the assertions of Theorem 1 are not valid

∞ 
∞
if the series an and bn are not convergent.
n=0 n=0
−n
solution Let an = 2 − 2n and bn = 2n . Then, both

∞ 
∞
an and bn
n=0 n=0

diverge, so the sum


∞ 
∞
an + bn
n=0 n=0

is not defined. However,


∞ 
∞ 
∞
(an + bn ) = ((2−n − 2n ) + 2n ) = 2−n = 1
n=0 n=0 n=0

Further Insights and Challenges

In Exercises 67–69, use the formula
1 − rN
1 + r + r2 + · · · + r N−1 = 6
1−r
67. Professor George Andrews of Pennsylvania State University observed that we can use Eq. (6) to calculate
the derivative of f (x) = xN (for N ≥ 0). Assume that a  0 and let x = ra. Show that
xN − aN rN − 1
f  (a) = lim = aN−1 lim
x→a x−a r→1 r − 1

and evaluate the limit.

solution According to the definition of derivative of f (x) at x = a

xN − aN
f  (a) = lim
x→a x − a
48 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

Now, let x = ra. Then x → a if and only if r → 1, and


x N
− a N
(ra) N
− a N aN r N − 1 rN − 1
f  (a) = lim = lim = lim = aN−1 lim
x→a x − a r→1 ra − a r→1 a (r − 1) r→1 r − 1

By formula (6) for a geometric sum,

1 − rN rN − 1
= = 1 + r + r2 + · · · + r N−1
1−r r−1
so
rN − 1 
lim = lim 1 + r + r2 + · · · + r N−1 = 1 + 1 + 12 + · · · + 1N−1 = N
r→1 r − 1 r→1

Therefore, f  (a) = aN−1 · N = NaN−1

68. Pierre de Fermat used geometric series to compute the area under the graph of f (x) = xN over [0, A]. For
0 < r < 1, let F(r) be the sum of the areas of the infinitely many right-endpoint rectangles with endpoints
Arn , as in Figure 7. As r tends to 1, the rectangles become narrower and F(r) tends to the area under the
graph.
1−r
(a) Show that F(r) = AN+1 .
1 − r N+1
A
(b) Use Eq. (6) to evaluate xN dx = lim F(r).
0 r→1

f (x) = x N

x
r 3A r 2A rA A
FIGURE 7

solution
(a) The area of the rectangle whose base extends from x = rn A to x = rn−1 A is

(rn−1 A)N (rn−1 A − rn A)

Hence, F(r) is the sum


∞ 
N  
∞ 
∞
F(r) = rn−1 A rn−1 A − rn A = r(n−1)N rn−1 (1 − r)AN+1 = AN+1 (1 − r) rnN−N+n−1
n=1 n=1 n=1

AN+1 (1 − r) 
∞ 
n AN+1 (1 − r) r N+1 1−r
= r N+1 = · = AN+1
r N+1 n=1
r N+1 1 − r N+1 1 − r N+1

(b) Using the result from part (a) and Eq. (6) from Exercise 67,
1−r 1 1 AN+1 A
lim F(r) = AN+1 lim = AN+1 lim = AN+1 · = = xN dx
r→1 r→1 1 − r N+1 r→1 1 + r + r + · · · + r
2 N N+1 N+1 0

69. Verify the Gregory–Leibniz formula in part (d) as follows.

(a) Set r = −x2 in Eq. (6) and rearrange to show that
1 (−1)N x2N
= 1 − x 2
+ x 4
− · · · + (−1) N−1 2N−2
x +
1 + x2 1 + x2
(b) Show, by integrating over [0, 1], that

π 1 1 1 (−1)N−1 1
x2N dx
= 1 − + − + ··· + + (−1)N
4 3 5 7 2N − 1 0 1 + x2
 S E C T I O N 10.2 Summing an Infinite Series (LT SECTION 11.2) 49

(c) Use the Comparison Theorem for integrals to prove that

1
x2N dx 1
0≤ ≤
0 1+x 2 2N + 1
Hint: Observe that the integrand is ≤ x2N .
(d) Prove that
π 1 1 1 1
= 1 − + − + − ···
4 3 5 7 9
 
Hint: Use (b) and (c) to show that the partial sums S N satisfy S N − π4  ≤ 1
2N+1 , and thereby conclude that
lim S N = π4 .
N→∞

solution
(a) Start with Eq. (6), and substitute −x2 for r:
1 − rN
1 + r + r2 + · · · + r N−1 =
1−r
1 − (−1)N x2N
1 − x2 + x4 + · · · + (−1)N−1 x2N−2 =
1 − (−x2 )
1 (−1)N x2N
1 − x2 + x4 + · · · + (−1)N−1 x2N−2 = −
1 + x2 1 + x2
1 (−1)N x2N
= 1 − x 2
+ x 4
+ · · · + (−1) N−1 2N−2
x +
1 + x2 1 + x2
(b) The integrals of both sides must be equal. Now,
1
1
1 −1  π
dx = tan x  = tan−1 1 − tan−1 0 =
0 1+x 2
0 4
while
1  
(−1)N x2N
1 − x2 + x4 + · · · + (−1)N−1 x2N−2 + dx
0 1 + x2
  1
1 1 1 x2N dx
= x − x3 + x5 + · · · + (−1)N−1 x2N−1 + (−1)N
3 5 2N − 1 0 1 + x2
1
1 1 1 x2N dx
=1− + + · · · + (−1)N−1 + (−1)N
3 5 2N − 1 0 1 + x2
(c) Note that for x ∈ [0, 1], we have 1 + x2 ≥ 1, so that
x2N
0≤ ≤ x2N
1 + x2
By the Comparison Theorem for integrals, we then see that
1 1 1
x2N dx
x2N+1  =
1 1
0≤ ≤ x2N dx =
0 1 + x2 0 2N + 1 0 2N + 1
(d) Write
1
an = (−1)n , n≥1
2n − 1
and let S N be the partial sums. Then
   
S − π  = (−1)N
1
x2N dx  1
x2N dx 1
 N 4   = ≤
0 1 + x2  0 1+x 2 2N + 1
π
Thus limN→∞ S N = so that
4
π 1 1 1 1
= 1 − + − + − ...
4 3 5 7 9
50 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

70. Cantor’s Disappearing Table (following Larry Knop of Hamilton College) Take a table of length L
(Figure 8). At Stage 1, remove the section of length L/4 centered at the midpoint. Two sections remain, each
with length less than L/2. At Stage 2, remove sections of length L/42 from each of these two sections (this
stage removes L/8 of the table). Now four sections remain, each of length less than L/4. At Stage 3, remove
the four central sections of length L/43 , and so on.
(a) Show that at the Nth stage, each remaining section has length less than L/2N and that the total amount
of table removed is
 
1 1 1 1
L + + + · · · + N+1
4 8 16 2
(b) Show that in the limit as N → ∞, precisely one-half of the table remains.
This result is intriguing, because there are no nonzero intervals of table left (at each stage, the remaining
sections have a length less than L/2N ). So, the table has “disappeared.” However, we can place any object
longer than L/4 on the table. The object will not fall through because it will not fit through any of the removed
sections.
L/16 L/4 L/16

FIGURE 8

solution
(a) After the Nth stage, the total amount of table that has been removed is
   
L 2L 4L 2N−1 L 1 1 1 2N−1 1 1 1 1
+ 2 + 3 + ··· + = L + + + · · · + = L + + + · · · +
4 4 4 4N 4 8 16 22N 4 8 16 2N+1
At the first stage (N = 1), there are two remaining sections each of length
L− L
3L L
4
= <
2 8 2
L
Suppose that at the Kth stage, each of the 2K remaining sections has length less than K . The (K + 1)st stage
2
L
is obtained by removing the section of length K+1 centered at the midpoint of each segment in the Kth stage.
4
Let ak and aK+1 , respectively, denote the length of each segment in the Kth and (K + 1)st stage. Then,
⎛ ⎞
aK − 4K+1 − 4K+1 L ⎜⎜⎜⎜ 1 − 2K+2 ⎟⎟⎟⎟
L L L 1
2K L 1 L
aK+1 = < = K ⎜⎝ ⎟⎠ < K · = K+1
2 2 2 2 2 2 2
L
Thus, by mathematical induction, each remaining section at the Nth stage has length less than .
2N
(b) From part (a), we know that after N stages, the amount of the table that has been removed is
   N
1 1 1 1 1
L + + + · · · + N+1 = n+1
4 8 16 2 n=1
2

As N → ∞, the amount of the table that has been removed becomes a geometric series whose sum is
∞  n 1
1 1 1
L =L 4 1 = L
n=1
2 2 1− 2 2

Thus, the amount of table that remains is L − 12 L = 12 L.

71. The Koch snowflake (described in 1904 by Swedish mathematician Helge von Koch) is an infinitely
jagged “fractal” curve obtained as a limit of polygonal curves (it is continuous but has no tangent line at any
point). Begin with an equilateral triangle (Stage 0) and produce Stage 1 by replacing each edge with four
edges of one-third the length, arranged as in Figure 9. Continue the process: At the nth stage, replace each
edge with four edges of one-third the length of the edge from the (n − 1)st stage.
 S E C T I O N 10.3 Convergence of Series with Positive Terms (LT SECTION 11.3) 51

(a) Show that the perimeter Pn of the polygon at the nth stage satisfies Pn = 43 Pn−1 . Prove that lim Pn = ∞.
n→∞
The snowflake has infinite length.
(b) Let A0 be the area of the original equilateral triangle. Show that (3)4n−1 new triangles are added at the
nth stage, each with area A0 /9n (for n ≥ 1). Show that the total area of the Koch snowflake is 85 A0 .

Stage 1 Stage 2 Stage 3

FIGURE 9

solution
(a) Each edge of the polygon at the (n − 1)st stage is replaced by four edges of one-third the length; hence
the perimeter of the polygon at the nth stage is 43 times the perimeter of the polygon at the (n − 1)th stage.
That is, Pn = 43 Pn−1 . Thus,
 2  3
4 4 4 4 4
P1 = P0 ; P2 = P1 = P0 , P3 = P2 = P0
3 3 3 3 3
4 n
and, in general, Pn = 3 P0 . As n → ∞, it follows that
 n
4
lim Pn = P0 lim =∞
n→∞ n→∞ 3

(b) When each edge is replaced by four edges of one-third the length, one new triangle is created. At the
(n − 1)st stage, there are 3 · 4n−1 edges in the snowflake, so 3 · 4n−1 new triangles are generated at the nth
stage. Because the area of an equilateral triangle is proportional to the square of its side length and the side
length for each new triangle is one-third the side length of triangles from the previous stage, it follows that
the area of the triangles added at each stage is reduced by a factor of 19 from the area of the triangles added
at the previous stage. Thus, each triangle added at the nth stage has an area of A0 /9n . This means that the nth
stage contributes
 n
A0 3 4
3 · 4n−1 · n = A0
9 4 9
to the area of the snowflake. The total area is therefore
∞  n
3  4
4
3 3 4 8
A = A0 + A0 = A0 + A0 9 = A0 + A0 · = A0
4 n=1 9 4 1− 4
9
4 5 5

10.3 Convergence of Series with Positive Terms

(LT Section 11.3)
Preliminary Questions

∞
1. For the series an , if the partial sums S N are increasing, then (choose the correct conclusion)
n=1
(a) {an } is an increasing sequence.
(b) {an } is a positive sequence.
solution The correct response is (b). Recall that S N = a1 + a2 + a3 + · · · + aN ; thus, S N − S N−1 = aN . If
S N is increasing, then S N − S N−1 ≥ 0. It then follows that aN ≥ 0; that is, {an } is a positive sequence.
2. What are the hypotheses of the Integral Test?
solution The hypotheses for the Integral Test are: A function f (x) such that an = f (n) must be positive,
decreasing, and continuous for x ≥ 1.
52 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)


∞
3. Which test would you use to determine whether n−3.2 converges?
n=1

solution Because n−3.2 = 1

n3.2
, we see that the indicated series is a p-series with p = 3.2 > 1. Therefore, the
series converges.

∞
1
4. Which test would you use to determine whether √
converges? n=1
2n + n

solution Because
 n
1 1 1
√ < n =
2 + n
n 2 2
and
∞  n
 1
n=1
2

is a convergent geometric series, the Direct Comparison Test would be an appropriate choice to establish that
the given series converges.
∞
e−n ∞
1
5. Ralph hopes to investigate the convergence of by comparing it with . Is Ralph on the right
n=1
n n=1
n
track?

solution No, Ralph is not on the right track. For n ≥ 1,

e−n 1
<
n n

∞
1
however, is a divergent series. The Direct Comparison Test therefore does not allow us to draw a
n=1
n

∞
e−n
conclusion about the convergence or divergence of the series .
n=1
n

Exercises
In Exercises 1–12, use the Integral Test to determine whether the infinite series is convergent.


∞
1
1.
n=1
(n + 1)4
1
solution Let f (x) = . This function is continuous, positive, and decreasing on the interval x ≥ 1,
(x + 1)4
so the Integral Test applies. Moreover,
∞ R  R   
dx
= lim
dx
= lim −
1  = lim 1 − 1
=
1
1 (x + 1)
4 R→∞ 1 (x + 1)4 R→∞ 3(x + 1)3 1 R→∞ 24 3(R + 1)3 24

∞
1
The integral converges; hence, the series also converges.
n=1
(n + 1)4

∞
1
2.
n=1
n+3
1
solution Let f (x) = . This function is continuous, positive and decreasing on the interval x ≥ 1, so
x+3
the Integral Test applies. Moreover,
∞ R
dx dx
= lim = lim (ln(R + 3) − ln 4) = ∞
1 x + 3 R→∞ 1 x + 3 R→∞

∞
1
The integral diverges; hence, the series also diverges.
n=1
n+3
 S E C T I O N 10.3 Convergence of Series with Positive Terms (LT SECTION 11.3) 53


∞
3. n−1/3
n=1

1
Let f (x) = x− 3 = √3 . This function is continuous, positive, and decreasing on the interval x ≥ 1,
1
solution
x
so the Integral Test applies. Moreover,
∞ R
3 
x−1/3 dx = lim x−1/3 dx = lim R2/3 − 1 = ∞
1 R→∞ 1 2 R→∞

∞
The integral diverges; hence, the series n−1/3 also diverges.
n=1

∞
1
4. √
n=5 n−4
1
solution Let f (x) = √. This function is continuous, positive and decreasing on the interval x ≥ 5, so
x−4
the Integral Test applies. Moreover,
∞
dx R
dx √
√ = lim √ = 2 lim R−4−1 =∞
5 x−4 R→∞ 5 x−4 R→∞


∞
1
The integral diverges; hence, the series √ also diverges.
n=5 n−4

∞
n2
5.
n=25
(n3 + 9)5/2

x2
solution Let f (x) = 5/2
. This function is positive and continuous for x ≥ 25. Moreover, because
x3 + 9
5/2 3/2

2x(x3 + 9) − x2 · 52 (x3 + 9) · 3x2 x(36 − 11x3 )
f (x) = =
(x3 + 9)5 2(x3 + 9)7/2
we see that f  (x) < 0 for x ≥ 25, so f is decreasing on the interval x ≥ 25. The Integral Test therefore applies.
To evaluate the improper integral, we use the substitution u = x3 + 9, du = 3x2 dx. We then find
∞ R R +9 3
x2 x2 1 du
dx = lim dx = lim
25 (x3 + 9)5/2 R→∞ 25 (x 3 + 9)5/2 3 R→∞ 15634 u5/2
 
2 1 1 2
= − lim − =
9 R→∞ (R3 + 9)3/2 156343/2 9 · 156343/2

∞
n2
The integral converges; hence, the series 5/2
also converges.
n=25 n3 + 9

∞
n
6.
n=1
(n2 + 1)3/5
x
solution Let f (x) = . Because
(x2 + 1)3/5

(x2 + 1)3/5 − x · 65 x(x2 + 1)−2/5 1 − 15 x2

f  (x) = =
(x2 + 1)6/5 (x2 + 1)8/5
√
we see that f  (x) < 0 for x > 5 ≈ 2.236. We conclude that f is decreasing on the interval x ≥ 3. Since
f is also positive and continuous on this interval, the Integral Test can be applied. To evaluate the improper
integral, we make the substitution u = x2 + 1, du = 2x dx. This gives
∞
x R
x 1 R2 +1
du 5 
dx = lim dx = lim = lim (R2 + 1)2/5 − 102/5 = ∞
3 (x2 + 1)3/5 R→∞ 3 (x2 + 1)3/5 2 R→∞ 10 u3/5 4 R→∞
54 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

n 
∞
The integral diverges; therefore, the series also diverges. Since the divergence of the series is
n=3
+ 1)3/5 (n2
2
n ∞
n
not affected by adding the finite sum , the series also diverges.
n=1
(n + 1)
2 3/5
n=1
(n + 1)3/5
2


∞
1
7.
n=1
n2 +1
1
solution Let f (x) = 2 . This function is positive, decreasing and continuous on the interval x ≥ 1,
x +1
hence the Integral Test applies. Moreover,
∞
dx R
dx  π π π π
−1
= lim = lim tan R − = − =
1 x + 1 R→∞
2
1 x2 + 1 R→∞ 4 2 4 4

∞
1
The integral converges; hence, the series also converges.
n=1
n2 +1

∞
1
8.
n=4
n2 − 1
1
solution Let f (x) = 2 .This function is continuous, positive and decreasing on the interval x ≥ 4;
x −1
therefore, the Integral Test applies. We compute the improper integral using partial fractions:
∞ R ⎛ ⎞   
⎜⎜⎜ ⎟⎟⎟ x − 1 R 1
1 1
dx 1 R−1 3
= lim ⎜⎝ 2
− ⎟⎠ dx = lim ln
2
 = lim ln − ln
x2 − 1 R→∞ x−1 x+1 2 R→∞ x + 1 4 2 R→∞ R+1 5
4 4
 
1 3 1 3
= ln 1 − ln = − ln
2 5 2 5

∞
1
The integral converges; hence, the series also converges.
n=4
n2 −1

∞
1
9.
n=1
n(n + 5)

1
solution Let f (x) = . This function is positive, continuous, and decreasing on the interval x ≥ 1,
x(x + 5)
so the Integral Test applies. We compute the improper integral using partial fractions:
R    
∞
dx 1 1 1 1 x R 1 R 1
= lim − dx = lim ln  = lim ln − ln
1 x(x + 5) 5 R→∞ 1 x x + 5 5 R→∞ x + 5 1 5 R→∞ R+5 6
 
1 1 1
= ln 1 − ln = ln 6
5 6 5

∞
1
The integral converges; hence, the series converges.
n=1
n(n + 5)

∞
ne−n
2
10.
n=1

Let f (x) = xe−x . This function is continuous and positive on the interval x ≥ 1. Moreover, be-
2
solution
cause

f  (x) = 1 · e−x + x · e−x · (−2x) = e−x 1 − 2x2
2 2 2

we see that f  (x) < 0 for x ≥ 1, so f is decreasing on this interval. To compute the improper integral we make
the substitution u = x2 , du = 2x dx. Then, we find
 S E C T I O N 10.3 Convergence of Series with Positive Terms (LT SECTION 11.3) 55

∞ R
1 R2
1  2 1
xe−x dx = lim xe−x dx = e−u du = − lim e−R − e−1 =
2 2

1 R→∞ 1 2 1 2 R→∞ 2e

∞
ne−n also converges.
2
The integral converges; hence, the series
n=1

∞
1
11.
n=2
n(ln n)3/2
1
solution Let f (x) = . This function is positive and continuous for x ≥ 2. Moreover,
x(ln x)3/2
 
 1 3 1/2 1 3 + 2 ln x
f (x) = − 2 1 · (ln x) + x · (ln x) ·
3/2
=− 2 (1)
x (ln x)3 2 x 2x (ln x)5/2
Since ln x > 0 for x > 1, f  (x) is negative for x > 1; hence, f is decreasing for x ≥ 2. To compute the improper
1
integral, we make the substitution u = ln x, du = dx. We obtain:
x
∞ R ln R
1 1 du
dx = lim dx = lim
2 x(ln x)3/2 R→∞ 2 x(ln x)
3/2 R→∞ ln 2 u
3/2
ln R
= − lim 2u−1/2 
R→∞ u=ln 2

= −2 lim (ln R)−1/2 − (ln 2)−1/2 = 2(ln 2)−1/2
R→∞


∞
1
The integral converges; hence, the series also converges.
n=2
n(ln n)3/2

∞
ln n
12.
n=1
n2
ln x
solution Let f (x) = . Because
x2
1
· x2 − 2x ln x x (1 − 2 ln x) 1 − 2 ln x
f  (x) = x
= =
x4 x4 x3
√
we see that f  (x) < 0 for x > e ≈ 1.65. We conclude that f is decreasing on the interval x ≥ 2. Since f is
also positive and continuous on this interval, the Integral Test can be applied. By Integration by Parts, we
find
ln x ln x ln x 1
dx = − + x−2 dx = − − +C
x2 x x x
therefore,
∞  
ln x R
ln x 1 ln 2 1 ln R 1 + ln 2 ln R
dx = lim dx = lim + − − = − lim
2 x2 R→∞ 2 x 2 R→∞ 2 2 R R 2 R→∞ R
We compute the resulting limit using L’Hôpital’s Rule:
ln R 1/R
lim = lim =0
R→∞ R R→∞ 1

Hence,
∞
ln x 1 + ln 2
2
dx =
2 x 2

∞
ln n
The integral converges; therefore, the series also converges. Since the convergence of the series is
n=2
n2

1
ln n 
∞
ln n
not affected by adding the finite sum , the series also converges.
n=1
n2 n=1
n2
56 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)


∞
1  ∞
13. Show that converges by using the Direct Comparison Test with n−3 .
n=1
n + 8n
3
n=1

∞
solution We compare the series with the p-series n−3 . For n ≥ 1,
n=1

1 1
≤
n3 + 8n n3
 ∞
1 ∞
1
Since 3
converges (it is a p-series with p = 3 > 1), the series 3 + 8n
also converges by the Direct
n=1
n n=1
n
Comparison Test.
 ∞
1 ∞
14. Show that √ diverges by comparing with n−1 .
n=2 n −3
2
n=2

solution For n ≥ 2,
1 1 1
√ ≥ √ =
n −3
2 n 2 n

∞
1 
∞
1
The harmonic series diverges, and it still diverges if we drop the first term. Thus, the series also
n=1
n n=2
n

∞
1
diverges. The Direct Comparison Test now lets us conclude that the larger series √ also diverges.
n=2 n −3
2


∞
1
15. For √ , verify that for n ≥ 1,
n=1
n+ n

1 1 1 1
√ ≤ , √ ≤ √
n+ n n n+ n n
1 1
Can either inequality be used to show that the series diverges? Show that √ ≥ for n ≥ 1 and
n + n 2n
conclude that the series diverges.
√ √ √
solution For n ≥ 1, n + n ≥ n and n + n ≥ n. Taking the reciprocal of each of these inequalities
yields
1 1 1 1
√ ≤ and √ ≤ √
n+ n n n+ n n

∞
1 1 ∞  1 ∞ 1 ∞
These inequalities indicate that the series √ is smaller than both and √ ; however,
n=1
n+ n n=1
n n=1
n n=1
n
∞
1
and √ both diverge so neither inequality allows us to show that the original series diverges.
n
n=1 √ √
On the other hand, for n ≥ 1, n ≥ n, so 2n ≥ n + n and
1 1
√ ≥
n + n 2n
 ∞
1 ∞
1
The series =2 diverges, since the harmonic series diverges. The Direct Comparison Test then
n=1
2n n=1
n
∞
1
lets us conclude that the larger series √ also diverges.
n=1
n+ n

∞
1
16. Which of the following inequalities can be used to study the convergence of √ ? Explain.
n=2
n2 + n

1 1 1 1
√ ≤ √ , √ ≤ 2
n2 + n n n2 + n n
 S E C T I O N 10.3 Convergence of Series with Positive Terms (LT SECTION 11.3) 57

∞
1 ∞
1
solution The series √ is a divergent p-series, hence the series √ also diverges. The first in-
n=1
n n=2
n
∞
1
equality given above therefore establishes that 2+
√ is smaller than a divergent series, which does not
n=2
n n
 ∞
1
allow us to conclude whether √ converges or diverges.
n=2
n + n
2

∞
1
However, the second inequality given above establishes that 2+
√ is smaller than the convergent
n=2
n n
∞
1  ∞
1
p-series . By the Direct Comparison Test, we therefore conclude that √ also converges.
n=2
n2
n=2
n + n
2

In Exercises 17–28, use the Direct Comparison Test to determine whether the infinite series is convergent.
∞
1
17. n
n=1
n2
∞  n
 1
solution We compare with the geometric series . For n ≥ 1,
n=1
2
 n
1 1 1
≤ =
n2n 2n 2
∞  n
 1
Since converges (it is a geometric series with r = 12 ), we conclude by the Direct Comparison Test
n=1
2
∞
1
that n
also converges.
n=1
n2

∞
n3
18.
n=1
n5 + 4n + 1
solution For n ≥ 1,

n3 n3 1
≤ = 2
n + 4n + 1 n
5 5 n
 ∞
1
The series is a p-series with p = 2 > 1, so it converges. By the Direct Comparison Test we can
n=1
n2
∞
n3
therefore conclude that the series also converges.
n=1
n5 + 4n + 1

∞
1
19.
n=1
n1/3 + 2n
solution For n ≥ 1,
1 1
≤
n1/3 + 2n 2n
# 1 1
The series ∞ n=1 n is a geometric series with r = , so it converges. By the Comparison test, so does
2 2
∞
1
.
n=1
n + 2n
1/3


∞
1
20. √
n=1 n3 + 2n − 1
solution For n ≥ 1, we have 2n − 1 ≥ 0 so that
1 1 1
√ ≤ √ = 3/2
n3 + 2n − 1 n3 n
58 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

This latter series is a p-series with p = 3

2 > 1, so it converges. By the Comparison Test, so does
∞
1
√ .
n=1 n + 2n − 1
3


∞
4
21.
m=1
m! + 4m

solution For m ≥ 1,
 m−1
4 4 1
≤ =
m! + 4m 4m 4
∞  m−1
 1 1
The series is a geometric series with r = , so it converges. By the Direct Comparison Test we
m=1
4 4

∞
4
can therefore conclude that the series also converges.
m=1
m! + 4m
∞ √
n
22.
n=4
n−3

solution For n ≥ 4,
√ √
n n 1
≥ = 1/2
n−3 n n
 ∞
1 1
The series 1/2
is a p-series with p = < 1, so it diverges, and it continues to diverge if we drop the
n=1
n 2
 ∞
1
terms n = 1, 2, 3; that is, 1/2
also diverges. By the Direct Comparison Test we can therefore conclude
n
∞ √ n=4
n
that series diverges.
n=4
n − 3

∞
sin2 k
23.
k=1
k2

solution For k ≥ 1, 0 ≤ sin2 k ≤ 1, so

sin2 k 1
0≤ 2
≤ 2
k k
 ∞
1
The series 2
is a p-series with p = 2 > 1, so it converges. By the Direct Comparison Test we can
k=1
k
∞
sin2 k
therefore conclude that the series also converges.
k=1
k2

∞
k2/9
24.
k=2
k10/9 − 1

solution We have for k ≥ 2

k2/9 k2/9 1
> 10/9 = 8/9
k10/9−1 k k
 ∞
1
The series 8/9
is a p-series with p = 89 < 1, so it diverges. By the Direct Comparison Test we can
k=1
k
∞
k2/9
therefore conclude that the series also diverges.
k=2
k 10/9 −1
 S E C T I O N 10.3 Convergence of Series with Positive Terms (LT SECTION 11.3) 59


∞
2
25.
n=1
3n + 3−n

solution Since 3−n > 0 for all n,

 n
2 2 1
−n
≤ n =2
3 +3
n 3 3

∞  n
1 1
The series 2 is a geometric series with r = , so it converges. By the Direct Comparison Test we can
n=1
3 3
∞
2
therefore conclude that the series also converges.
n=1
3 + 3−n
n


∞
2−k
2
26.
k=1

solution For k ≥ 1, k2 ≥ k and

 k
1 1 1
k 2
≤ k
=
2 2 2

∞  k
 1 1
The series is a geometric series with r = , so it converges. By the Direct Comparison Test we can
k=1
2 2
∞
1 ∞
2−k also converges.
2
therefore conclude that the series k2
=
k=1
2 k=1


∞
1
27.
n=1
(n + 1)!

solution Note that for n ≥ 2,

(n + 1)! = 1 · 2 · 3 · · · n · (n + 1) ≤ 2n

n factors

so that

∞
1  1
∞  1 ∞
=1+ ≤1+
n=1
(n + 1)! n=2
(n + 1)! n=2
2n

#∞ 1 1 ∞
1
But is a geometric series with ratio r = , so it converges. By the comparison test,
n=2 n
2 2 n=1
(n + 1)!
converges as well.
∞
n!
28.
n=1
n3

solution Note that for n ≥ 4, we have (n − 1)(n − 2) > n [to see this, solve the equation (n − 1)(n − 2) = n
√
for n; the positive root is 2 + 2 ≈ 3.4]. Thus

∞
n!  n(n − 1)(n − 2)(n − 3)!  (n − 3)!  1
∞ ∞ ∞
= ≥ ≥
n=4
n3 n=4 n3 n=4
n n=4
n

# 1 #
But ∞ n=4 is the harmonic series, which diverges, so that ∞ n!
n=4 n3 also diverges. Adding back in the terms
n
for n = 1, 2, and 3 does not affect this result. Thus the original series diverges.
60 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

Exercise 29–34: For all a > 0 and b > 1, the inequalities ln n ≤ na , na < bn are true for n sufficiently large
(this can be proved using L’Hôpital’s Rule). Use this, together with the Direct Comparison Test, to determine
whether the series converges or diverges.

∞
ln n
29.
n=1
n3
solution For n sufficiently large (say n = k, although in this case n = 1 suffices), we have ln n ≤ n, so that

∞ ∞
ln n n ∞
1
3
≤3
= 2
n=k
n n=k
n n=k
n
# ln n
This is a p-series with p = 2 > 1, so it converges. Thus ∞ n=k n3 also converges; adding back in the finite
number of terms for 1 ≤ n ≤ k does not affect this result.
∞
1
30.
m=2
ln m
solution For m > 1 sufficiently large (say m = k, although in this case m = 2 suffices), we have ln m ≤ m,
so that
∞
1 ∞
1
≥
m=k
ln m m=k
m
This is the harmonic series, which diverges (the absence of the finite number of terms for m = 1, . . . , k − 1
∞
1
does not affect convergence). By the comparison theorem, also diverges (again, ignoring the finite
m=2
ln m
number of terms for m = 1, . . . , k − 1 does not affect convergence).
∞
(ln n)100
31.
n=1
n1.1
solution Choose N so that ln n ≤ n0.0005 for n ≥ N. Then also for n > N, (ln n)100 ≤ (n0.0005 )100 = n0.05 .
Then

∞
(ln n)100 
∞
n0.05 ∞
1
≤ =
n=N
n1.1 n=N
n1.1 n=N
n1.05

 ∞
1 #
But 1.05
is a p-series with p = 1.05 > 1, so is convergent. It follows that ∞n=N
(ln n)1 00
n1.1
is also convergent;
n=N
n
∞
(ln n)100
adding back in the finite number of terms for n = 1, 2, . . . , N − 1 shows that converges as well.
n=1
n1.1
 ∞
1
32.
n=1
(ln n)10
solution Choose N such that ln n ≤ n0.1 for n ≥ N; then also (ln n)10 ≤ n for n ≥ N. So we have

∞
1 1 ∞
≥
n=N
(ln n)10 n=N n
The latter sum is the harmonic series, which diverges, so the series on the left diverges as well. Adding back
∞
1
in the finite number of terms for n < N shows that diverges.
n=1
(ln n)10
∞
n
33. n
n=1
3
solution Choose N such that n ≤ 2n for n ≥ N. Then
∞ ∞  n
n 2
n
≤
n=N
3 n=N
3
2
The latter sum is a geometric series with r = < 1, so it converges. Thus the series on the left converges as
3
∞
n
well. Adding back in the finite number of terms for n < N shows that converges.
n=1
3n
 S E C T I O N 10.3 Convergence of Series with Positive Terms (LT SECTION 11.3) 61


∞
n5
34.
n=1
2n
solution Choose N such that n5 ≤ 1.5n for n ≥ N. Then
 ∞  n
n5  1.5
∞
≤
n=N
2n n=N 2

1.5
The latter sum is a geometric series with r = < 1, so it converges. Thus the series on the left converges
2
∞
n5
as well. Adding back in the finite number of terms for n < N shows that converges.
n=1
2n

∞
1
35. Show that sin converges. Hint: Use sin x ≤ x for x ≥ 0.
n=1
n2
solution For n ≥ 1,
1
0≤ ≤1<π
n2
therefore, sin n12 > 0 for n ≥ 1. Moreover, for n ≥ 1,
1 1
sin 2
≤ 2
n n
∞
1
The series is a p-series with p = 2 > 1, so it converges. By the Direct Comparison Test we can
n=1
n2
∞
1
therefore conclude that the series sin 2 also converges.
n=1
n

∞
sin(1/n)
36. Does converge? Hint: By Theorem 3 in Section 2.6, sin(1/n) > (cos(1/n))/n. Thus,
n=2
ln n
sin(1/n) > 1/(2n) for n > 2 [because cos(1/n) > 12 ].
solution No, it diverges. Either the Comparison Theorem or the Limit Comparison Theorem may be used.
Using the Comparison Theorem, recall that
sin x
> cos x for x > 0
x
so that sin x > x cos x. Substituting 1/n for x gives
   
1 1 1 cos(1/n) 1
sin > cos = ≥
n n n n 2n
 
1 1
since cos ≥ for n ≥ 2. Thus
n 2

∞
sin(1/n) 
∞
1
>
n=1
ln n n=1
2n ln n

Apply the Integral Test to the latter expression, making the substitution u = ln x:
∞ ∞ ∞
du = ln u0
1 1 1 1
dx =
1 2x ln x 2 0 u 2
and the integral diverges. Thus

∞
1 
∞
sin(1/n)
diverges, and thus diverges as well.
n=1
2n ln n n=1
ln n

Applying the Limit Comparison Test is similar but perhaps simpler: Recall that
sin(1/x) sin x
lim = lim =1
x→∞ 1/x x→0 x
62 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

1/x
so apply the Limit Comparison Test with bn = :
ln x
sin(1/x) ln x sin(1/x)
L = lim · = lim =1
x→∞ ln x 1/x x→∞ 1/x
so that either both series converge or both diverge. But by the Integral Test as above,

∞
(1/x) 
∞
1
=
n=1
ln x n=1
x ln x

diverges.

In Exercises 37–46, use the Limit Comparison Test to prove convergence or divergence of the infinite series.

∞
n2
37.
n=2
n4 − 1

n2 n2 n2 1
solution Let an = 4 . For large n, 4 ≈ 4 = 2 , so we apply the Limit Comparison Test with
n −1 n −1 n n
1
bn = 2 . We find
n
n2
an n4 −1 n4
L = lim = lim = lim =1
n→∞ bn n→∞ 1 n→∞ n4 − 1
n2

 ∞
1  ∞
1
The series 2
is a p-series with p = 2 > 1, so it converges; hence, 2
also converges. Because L
n=1
n n=2
n
∞
n2
exists, by the Limit Comparison Test we can conclude that the series converges.
n=2
n −1
4


∞
1
38.
n=2
n(n − 1)

1 1 1 1
solution Let an = = 2 . For large n, 2 ≈ 2 , so we apply the Limit Comparison Test
n(n − 1) n − n n −n n
1
with bn = . We find
n2
1
an n2 −n n2
L = lim = lim = lim =1
n→∞ bn n→∞ 1 n→∞ n2 − n
n2

 ∞
1  ∞
1
The series 2
is a p-series with p = 2 > 1, so it converges; hence, the series 2
also converges.
n=1
n n=2
n
∞
1
Because L exists, by the Limit Comparison Test we can conclude that the series 2−n
converges.
n=2
n


∞
n
39. √
n=2 n3 + 1

n n n 1
solution Let an = √ . For large n, √ ≈ √ = √ , so we apply the Limit Comparison test
n3 −1 n3 −1 n 3 n
1
with bn = √ . We find
n
√n
√
an n3 −1 n3
L = lim = lim = lim √ =1
n→∞ bn n→∞ √1 n→∞ n3 − 1
n
 S E C T I O N 10.3 Convergence of Series with Positive Terms (LT SECTION 11.3) 63

∞
1 ∞
1
The series √ is a p-series with p = 1
2 < 1, so it diverges; hence, √ also diverges. Because L > 0,
n=1
n n=2
n

∞
n
by the Limit Comparison Test we can conclude that the series √ diverges.
n=2 n −1
3


∞
n3
40. √
n=2 n7 + 2n2 + 1

solution Let an be the general term of our series. Observe that

n3 n−3 · n3 1
an = √ = √ = √
7 2 −3
n − 2n + 1 n · n − 2n + 1
7 2 n − 2n−4 + n−6
This suggests that apply the Limit Comparison Test, comparing our series with

∞ ∞
1
bn = 1/2
n=2 n=2
n

The ratio of the terms is

√
an 1 n 1
= √ · = √
bn −4
n − 2n + n−6 1 1 − 2n−5 + n−7
Hence
an 1
lim = lim √ =1
n→∞ bn n→∞ 1 − 2n−5 + n−7
∞
1
The p-series 1/2
diverges since p = 1/2 < 1. Therefore, our original series diverges.
n=2
n


∞
3n + 5
41.
n=3
n(n − 1)(n − 2)

3n + 5 3n + 5 3n 3
solution Let an = . For large n, ≈ 3 = 2 , so we apply the Limit Com-
n(n − 1)(n − 2) n(n − 1)(n − 2) n n
1
parison Test with bn = 2 . We find
n
3n+5
an n(n+1)(n+2) 3n3 + 5n2
L = lim = lim = lim =3
n→∞ bn n→∞ 1 n→∞ n(n + 1)(n + 2)
n2

∞
1 ∞
1
The series 2
is a p-series with p = 2 > 1, so it converges; hence, the series 2
also converges. Be-
n=1
n n=3
n
∞
3n + 5
cause L exists, by the Limit Comparison Test we can conclude that the series converges.
n=3
n(n − 1)(n − 2)

∞
en + n
42.
n=1
e2n − n2

solution Let

en + n en + n 1
an = = n = n
e −n
2n 2 (e − n)(e + n) e − n
n

For large n,
1 1
≈ n = e−n
en −n e
64 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

so we apply the Limit Comparison Test with bn = e−n . We find

1
an n −n en
L = lim = lim e −n = lim n =1
n→∞ bn n→∞ e n→∞ e − n


∞ ∞  n

−n 1
The series e = is a geometric series with r = 1
e < 1, so it converges. Because L exists, by the
n=1 n=1
e
∞
en + n
Limit Comparison Test we can conclude that the series also converges.
n=1
e2n − n2


∞
1
43. √
n=1
n + ln n

solution Let
1
an = √
n + ln n
√ √ 1
For large n, n + ln n ≈ n, so apply the Comparison Test with bn = √ . We find
n
√
an 1 n 1
L = lim = lim √ · = lim =1
n→∞ bn n→∞ n + ln n 1 n→∞ 1 + ln
√n
n

 ∞
1 1
The series √ is a p-series with p = < 1, so it diverges. Because L exists, the Limit Comparison Test
n=1
n 2
tells us the the original series also diverges.


∞
ln(n + 4)
44.
n=1
n5/2

solution For large n, ln(n + 4) ≈ ln n n, so apply the Limit Comparison Test with bn = n−3/2 . We find
an ln(n + 4) 3/2 ln(n + 4) 1/(n + 4) 1
L = lim = lim · n = lim = lim = lim = 0.
n→∞ bn n→∞ n5/2 n→∞ n n→∞ 1 n→∞ n+4
 3
Since bn is a p-series with p = > 1, it converges; since L = 0 this implies that the original series
2
converges as well.

∞   ∞
1
45. 1 − cos Hint: Compare with n−2 .
n=1
n n=1

1 1
solution Let an = 1 − cos , and apply the Limit Comparison Test with bn = 2 . We find
n n
an 1 − cos n 1
1 − cos 1
− x12 sin 1x 1 sin 1
L = lim = lim = lim x
= lim = lim 1 x
n→∞ bn n→∞ 1
n2
x→∞ 1
x2
x→∞ − x23 2 x→∞
x

As x → ∞, u = 1
x → 0, so
1
1 sin 1 sin u 1
L= lim x
= lim =
2 x→∞ 1x 2 u→0 u 2

 ∞
1
The series is a p-series with p = 2 > 1, so it converges. Because L exists, by the Limit Comparison
n2
∞  
n=1
 1
Test we can conclude that the series 1 − cos also converges.
n=1
n
 S E C T I O N 10.3 Convergence of Series with Positive Terms (LT SECTION 11.3) 65


∞
46. (1 − 2−1/n ) Hint: Compare with the harmonic series.
n=1

1
solution Let an = 1 − 2−1/n , and apply the Limit Comparison Test with bn = . We find
n
an 1 − 2−1/n 1 − 2−1/x − 2 (ln 2)2 1 −1/x 
L = lim = lim = lim = lim x = lim 2−1/x ln 2 = ln 2
n→∞ bn n→∞ 1
n
x→∞ 1
x
x→∞ − x12 x→∞


∞
1
The harmonic series diverges; because L > 0, we can conclude by the Limit Comparison Test that the
n=1
n

∞
series (1 − 2−1/n ) also diverges.
n=1

In Exercises 47–76, determine convergence or divergence using any method covered so far.

∞
1
47.
n=4
n2 − 9

1 1
solution Apply the Limit Comparison Test with an = and bn = 2 :
n2 − 9 n
1
an n2 −9 n2
L = lim = lim = lim =1
n→∞ bn n→∞ 1
n2
n→∞ n2 −9
∞
1  ∞
1
Since the p-series 2
converges, the series 2
also converges. Because L exists, by the Limit Compar-
n=1
n n=4
n
∞
1
ison Test we can conclude that the series converges.
n=4
n −9
2


∞
cos2 n
48.
n=1
n2

solution For all n ≥ 1, 0 ≤ cos2 n ≤ 1, so

cos2 n 1
0≤ ≤ 2
n2 n
∞
1
The series 2
is a convergent p-series; hence, by the Direct Comparison Test we can conclude that the
n=1
n

∞
cos2 n
series also converges.
n=1
n2


∞ √
n
49.
n=1
4n + 9
√
n 1
solution Apply the Limit Comparison Test with an = and bn = √ :
4n + 9 n
√
n
an 4n+9 n 1
L = lim = lim = lim =
n→∞ bn n→∞ √1 n→∞ 4n + 9 4
n

 ∞
1
The series √ is a divergent p-series. Because L > 0, by the Limit Comparison Test we can conclude
n
n=1
∞ √
n
that the series also diverges.
n=1
4n + 9
66 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)


∞
n − cos n
50.
n=1
n3

n − sin n 1
solution Apply the Limit Comparison Test with an = and bn = 2 :
n3 n

n−sin n  
an n3 sin n
L = lim = lim 1
= lim 1 − =1
n→∞ bn n→∞ n→∞ n
n2

∞
1
The series 2
is a convergent p-series. Because L exists, by the Limit Comparison Test we can conclude
n=1
n
∞
n − sin n
that the series also converges.
n=1
n3


∞
n2 − 1
51.
n=1
n5 + 1

solution Observe that

n2 − 1 n2 1
< =
n5 + 1 n5 n3

∞
1
for n ≥ 1. The series is a convergent p-series, so by the Direct Comparison Test we can conclude that
n=1
n3

∞
n2 − 1
the series also converges.
n=1
n5 + 1


∞
1
52.
n=1
n2 + sin n

1 1
solution Apply the Limit Comparison Test with an = and bn = 2 :
n2 + sin n n

1
an n2 +sin n 1
L = lim = lim = lim =1
n→∞ bn n→∞ 1
n2
n→∞ 1 + sin
n2
n

∞
1
The series 2
is a convergent p-series. Because L exists, by the Limit Comparison Test we can conclude
n=1
n
∞
1
that the series also converges.
n=1
n + sin n
2


∞
53. (4/5)−n
n=5

solution

∞  −n
 ∞  n

4 5
=
n=5
5 n=5
4

5
which is a geometric series starting at n = 5 with ratio r = > 1. Thus the series diverges.
4
 S E C T I O N 10.3 Convergence of Series with Positive Terms (LT SECTION 11.3) 67

∞
1
54.
n=1
3n2

2
solution Because n2 ≥ n for n ≥ 1, 3n ≥ 3n and
 n
1 1 1
n2
≤ n
=
3 3 3
∞  n
 1 1
The series is a geometric series with r = , so it converges. By the Direct Comparison Test we can
n=1
3 3
∞
1
therefore conclude that the series also converges.
n=1
3n2


∞
1
55.
n=2
n3/2 ln n

solution For n ≥ 3, ln n > 1, so n3/2 ln n > n3/2 and

1 1
<
n3/2 ln n n3/2
∞
1 ∞
1
The series 3/2
is a convergent p-series, so the series 3/2
also converges. By the Direct Comparison
n=1
n n=3
n
 ∞
1 ∞
1
Test we can therefore conclude that the series 3/2 ln n
converges. Hence, the series 3/2 ln n
also
n=3
n n=2
n
converges.


∞
(ln n)12
56.
n=2
n9/8

solution By the comment preceding Exercise 29, we can choose N so that for n ≥ N, we have ln n < n1/192 .
Then also for n ≥ N we have (ln n)12 < n12/192 = n1/16 . Then
∞
(ln n)12  n1/16  1
∞ ∞
≤ =
n=N
n9/8 n=N
n9/8 n=N
n17/16
which is a convergent p-series. Thus the series on the left converges as well; adding back in the finite number
∞
(ln n)12
of terms for n ≤ N shows that converges.
n=2
n9/8


∞
57. 41/k
k=1

solution
lim ak = lim 41/k = 20 = 1  0
k→∞ k→∞


∞
therefore, the series 41/k diverges by the Divergence Test.
k=1


∞
4n
58.
n=1
5n − 2n

4n 4n
solution Apply the Limit Comparison Test with an = and bn = :
5n − 2n 5n
4n
an 5n −2n 1
L = lim = lim = lim
n→∞ bn n→∞ 4n
5n
n→∞ 1 − 2n
5n
68 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

Now,
2n 2x 2
lim n
= lim x = lim x =0
n→∞ 5 x→∞ 5 x→∞ 5 ln 5

so
an 1
L = lim = =1
n→∞ bn 1 − 0
∞  n
 4
The series is a convergent geometric series. Because L exists, by the Limit Comparison Test we can
n=1
5

∞
4n
conclude that the series also converges.
n=1
5n − 2n


∞
1
59.
n=2
(ln n)4

solution By the comment preceding Exercise 29, we can choose N so that for n ≥ N, we have ln n < n1/8 ,
so that (ln n)4 < n1/2 . Then

∞
1  1
∞

4
>
n=N
(ln n) n=N
n1/2

which is a divergent p-series. Thus the series on the left diverges as well, and adding back in the finite number
∞
1
of terms for n < N does not affect the result. Thus diverges.
n=2
(ln n)4


∞
2n
60.
n=1
3n −n

2n 2n
solution Apply the Limit Comparison Test with an = and bn = :
3n − n 3n
2n
an 3n −n 1
L = lim = lim = lim
n→∞ bn n→∞ 2n
3n
n→∞ 1 − 3nn

Now,
n x 1
lim = lim = lim =0
n→∞ 3n x→∞ 3 x x→∞ 3 x ln 3
so
an 1
L = lim = =1
n→∞ bn 1 − 0
∞  n
 2
The series is a convergent geometric series. Because L exists, by the Limit Comparison Test we can
n=1
3

∞
2n
conclude that the series also converges.
n=1
3n−n


∞
1
61.
n=3
n ln n − n

solution For n ≥ 2, n ln n − n ≤ n ln n; therefore,

1 1
≥
n ln n − n n ln n
 S E C T I O N 10.3 Convergence of Series with Positive Terms (LT SECTION 11.3) 69

1
Now, let f (x) = . For x ≥ 2, this function is continuous, positive, and decreasing, so the Integral Test
x ln x
applies. Using the substitution u = ln x, du = 1x dx, we find

∞ R ln R
dx dx du
= lim = lim = lim (ln(ln R) − ln(ln 2)) = ∞
2 x ln x R→∞ 2 x ln x R→∞ ln 2 u R→∞


∞
1
The integral diverges; hence, the series also diverges. By the Direct Comparison Test we can
n=2
n ln n

∞
1
therefore conclude that the series diverges.
n=2
n ln n − n


∞
1
62.
n=3
n(ln n)2 − n

tan−1 n 1
solution Apply the Limit Comparison Test with an = 2
and bn = 2 :
n n
tan−1 n
an π
L = lim = lim n2
1
= lim tan−1 n =
n→∞ bn n→∞ n→∞ 4
n2

∞
1
The series 2
is a convergent p-series. Because L exists, by the Limit Comparison Test we can conclude
n=1
n
∞
tan−1 n
that the series also converges.
n=1
n2

∞
1
63. n
n=1
n

solution For n ≥ 2, nn ≥ 2n ; therefore,

 n
1 1 1
≤ =
nn 2n 2
∞  n
 ∞  n

1 1
The series is a convergent geometric series, so also converges. By the Direct Comparison
n=1
2 n=2
2
∞
1 ∞
1
Test we can therefore conclude that the series n
converges. Hence, the series n
converges.
n=2
n n=1
n


∞
n2 − 4n3/2
64.
n=1
n3

solution Let an = 1
n and bn = − n3/2
4
. Then

∞ ∞
n2 − 4n3/2
(an + bn ) =
n=1 n=1
n3

∞
an diverges since it is the harmonic series
n=1

∞
3
bn is a p-series with p = > 1, so converges
n=1
2
# # #
Since an diverges and bn converges, it follows that (an + bn ) diverges.
70 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)


∞
1 + (−1)n
65.
n=1
n

solution Let

1 + (−1)n
an =
n
Then
⎧
⎪
⎪
⎨0 n odd
an = ⎪
⎪
⎩2 = 1
n = 2k even
2k k
#∞
Therefore, {an } consists of 0s in the odd places and the harmonic series in the even places, so an is just
# i=1
the sum of the harmonic series, which diverges. Thus ∞ i=1 an diverges as well.


∞
2 + (−1)n
66.
n=1
n3/2

solution For n ≥ 1

2 + (−1)n 2 + 1 3
0< 3/2
≤ 3/2 = 3/2
n n n
∞
1 ∞
3
The series 3/2
is a convergent p-series; hence, the series 3/2
also converges. By the Direct Compar-
n=1
n n=1
n
∞
2 + (−1)n
ison Test we can therefore conclude that the series converges.
n=1
n3/2


∞
1
67. sin
n=1
n

1 1
solution Apply the Limit Comparison Test with an = sin and bn = :
n n
sin 1n sin u
L = lim 1
= lim =1
n→∞
n
u→0 u

where u = 1n . The harmonic series diverges. Because L > 0, by the Limit Comparison Test we can conclude
∞
1
that the series sin also diverges.
n=1
n

∞
sin(1/n)
68. √
n=1
n

sin(1/n) 1/n
solution Apply the Limit Comparison Test with an = √ and bn = √ :
n n
√
sin(1/n) n sin(1/n) sin u
L = lim √ · = lim = lim =1
n→∞ n 1/n n→∞ 1/n u→0 u
# #
so that an and bn either both converge or both diverge. But

∞ ∞
1/n  1
∞
bn = √ =
n=1 n=1
n n=1 n3/2

∞
sin(1/n)
is a convergent p-series. Thus √ converges as well.
n=1
n
 S E C T I O N 10.3 Convergence of Series with Positive Terms (LT SECTION 11.3) 71


∞
2n + 1
69.
n=1
4n

solution For n ≥ 3, 2n + 1 < 2n , so

 n
2n + 1 2n 1
n
< n =
4 4 2
∞  n ∞  n
1 1
The series is a convergent geometric series, so also converges. By the Direct Comparison
n=1
2 n=3
2
∞
2n + 1 ∞
2n + 1
Test we can therefore conclude that the series n
converges. Finally, the series converges.
n=3
4 n=1
4n

∞
1
70. √
n
n=3 e

1 1
solution Apply the Limit Comparison Test with an = √ and bn = 4 :
n4 − n n
1√
an n4 − n n4
L = lim = lim = lim √ =1
n→∞ bn n→∞ 1
n4
n→∞ n4 − n
∞
1  ∞
1
The series 4
is a convergent p-series, so also converges. Because L exists, by the Limit Compari-
n=1
n n=3
n4
∞
1
son Test we can conclude that the series 4−
√ converges.
n=3
n n


∞
ln n
71.
n=4
n2 − 3n

solution By the comment preceding Exercise 29, we can choose N ≥ 4 so that for n ≥ N, ln n < n1/2 . Then

∞
ln n  n1/2
∞  1
∞
≤ =
n=N
n − 3n n=N n − 3n n=N n − 3n1/2
2 2 3/2

1 1
To evaluate convergence of the latter series, let an = 3/2 and bn = 3/2 , and apply the Limit
n − 3n1/2 n
Comparison Test:
an 1 1
L = lim = lim · n3/2 = lim =0
bn n→∞ n3/2 − 3n1/2
n→∞ n→∞ 1 − 3n−1
# # # #
Thus an converges if bn does. But bn is a convergent p-series. Thus an converges and, by the
comparison test, so does the original series. Adding back in the finite number of terms for n < N does not
affect convergence.


∞  
1
72. ln 1 +
n=1
n
 
1 n+1
solution Note that ln 1 + = ln = ln(n + 1) − ln n, so that
n n
k  
1
Sk = ln 1 + = (ln(k + 1) − ln k) + (ln k − ln(k − 1)) + · · · + (ln 2 − ln 1) = ln(k + 1)
n=1
n
so that

∞  
1
ln 1 + = lim S k = lim ln(k + 1) = ∞
n=1
n k→∞ k→∞

Thus the series diverges.

72 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)


∞
1
73.
n=2
n1/2 ln n

solution By the comment preceding Exercise 29, we can choose N ≥ 2 so that for n ≥ N, ln n < n1/4 . Then


∞
1  1∞
>
n=N
n1/2 ln n n=N n3/4

which is a divergent p-series. Thus the original series diverges as well—as usual, adding back in the finite
number of terms for n < N does not affect convergence.


∞
1
74.
n=1
n3/2 − (ln n)4

solution Let

1 1
an = , bn =
n3/2 − ln4 n n3/2

and apply the Limit Comparison Test:

an n3/2 1
L = lim = lim = lim
n→∞ bn n→∞ n3/2 − ln4 n 4
n→∞
1 − ln3/2n n

But by the comment preceding Exercise 29, ln n, and thus ln4 n, are eventually smaller than any positive
ln4 n #
power of n, so for n sufficiently large, 3/2 is arbitrarily small. Thus L = 1 and an converges if and only if
n
# # ∞
1
bn does. But bn is a convergent p-series, so converges.
n=1 n
3/2 − ln4 n


∞
4n2 + 15n
75.
n=2
3n4 − 5n2 − 17

solution Apply the Limit Comparison Test with

4n2 + 15n 4n2 4

an = , bn = =
3n4 − 5n2 − 17 3n4 3n2
We have

an 4n2 + 15n 3n2 12n4 + 45n3 12 + 45/n

L = lim = lim 4 · = lim = lim =1
n→∞ bn n→∞ 3n − 5n − 17
2 4 n→∞ 12n4 − 20n2 − 68 n→∞ 12 − 20/n2 − 68/n4

#∞ 
∞
4n2 + 15n
Now, bn is a p-series with p = 2 > 1, so converges. Since L = 1, we see that
n=1
n=1
3n4 − 5n2 − 17
converges as well.


∞
n
76.
n=1
4−n + 5−n

solution Note that

n n4n
lim = lim 
n→∞ 4−n +5−n n
n→∞
1+ 4 5

This limit approaches ∞/1 = ∞, so the terms of the sequence do not tend to zero. Thus the series is divergent.
 S E C T I O N 10.3 Convergence of Series with Positive Terms (LT SECTION 11.3) 73


∞
1
77. For which a does converge?
n=2
n(ln n)a

1
solution First consider the case a > 0 but a  1. Let f (x) = . This function is continuous, positive,
x(ln x)a
and decreasing for x ≥ 2, so the Integral Test applies. Now,
∞ R ln R  
dx dx du 1 1 1
= lim = lim = lim −
2 x(ln x)a R→∞ 2 x(ln x)a R→∞ ln 2 ua 1 − a R→∞ (ln R)a−1 (ln 2)a−1

Because
⎧
1 ⎪
⎪∞,
⎨ 0<a<1
lim = ⎪
⎪0,
R→∞ (ln R)a−1 ⎩ a>1

we conclude the integral diverges when 0 < a < 1 and converges when a > 1. Therefore


∞
1
converges for a > 1 and diverges for 0 < a < 1
n=2
n(ln n)a

∞
1 1
Next, consider the case a = 1. The series becomes . Let f (x) = . For x ≥ 2, this function is
n=2
n ln n x ln x
continuous, positive, and decreasing, so the Integral Test applies. Using the substitution u = ln x, du = 1
x dx,
we find
∞ R ln R
dx dx du
= lim = lim = lim (ln(ln R) − ln(ln 2)) = ∞
2 x ln x R→∞ 2 x ln x R→∞ ln 2 u R→∞

The integral diverges; hence, the series also diverges.


∞
(ln n)b
Finally, consider the case a < 0. Let b = −a > 0 so the series becomes . Since ln n > 1 for all
n=2
n
n ≥ 3, it follows that

(ln n)b 1
(ln n)b > 1 so >
n n

∞
1 
∞
(ln n)b
The series diverges, so by the Direct Comparison Test we can conclude that also diverges.
n=3
n n=3
n
∞
(ln n)b
Consequently, diverges. Thus,
n=2
n


∞
1
diverges for a < 0
n=2
n(ln n)a

To summarize:

∞
1
converges if a > 1 and diverges if a ≤ 1
n=2
n(ln n)a


∞
1
78. For which a does converge?
n=2
na ln n

solution First consider the case a > 1. For n ≥ 3, ln n > 1 and

1 1
< a
na ln n n
74 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

∞
1 ∞
1
The series a
is a p-series with p = a > 1, so it converges; hence, also converges. By the Direct
n=1
n n=3
na
 ∞
1
Comparison Test we can therefore conclude that the series a ln n
converges, which implies the series
n=3
n
∞
1
a
also converges.
n=2
n ln n
For a ≤ 1, na ≤ n so
1 1
≥
na ln n n ln n
1
for n ≥ 2. Let f (x) = . For x ≥ 2, this function is continuous, positive, and decreasing, so the Integral
x ln x
Test applies. Using the substitution u = ln x, du = 1x dx, we find
∞ R ln R
dx dx du
= lim = lim = lim (ln(ln R) − ln(ln 2)) = ∞
2 x ln x R→∞ 2 x ln x R→∞ ln 2 u R→∞


∞
1
The integral diverges; hence, the series also diverges. By the Direct Comparison Test we can
n=2
n ln n

∞
1
therefore conclude that the series diverges.
n=2
na ln n
To summarize,

∞
1
converges for a > 1 and diverges for a ≤ 1.
n=2
na ln n


∞
n2
79. For which values of p does converge?
n=1
(n3 + 1) p

solution Using the Integral Test, we want to evaluate

∞
x2
dx
1 (x3 + 1) p
Use the substitution u = x + 1, so that
3 1
3 du = x2 dx. Then
∞ ∞
x2 1 1
dx = du
1 (x + 1)
3 p 3 2 up
But by the proof of Theorem 3 in the text, this integral converges for p > 1 and diverges for p ≤ 1. Thus the
#
series ∞ n2
n=1 (n3 +1) p converges for p > 1 and diverges for p ≤ 1.


∞
ex
80. For which values of p does converge?
n=1
(1 + e2x ) p

solution Using the Integral Test, we want to evaluate

∞
ex
dx
1 (1 + e2x ) p
Use the substitution u = e , so that du = e dx; then
x x

∞ ∞
ex 1
dx = du
1 (1 + e2x ) p e (1 + u2 ) p
Claim that this converges for p > and diverges otherwise. To see this, first assume p > 12 . Then 1 + u2 > u2 ,
1
 p 2
so that (1 + u2 ) p > u2 = u2p , so that
∞ ∞
1 1
du ≤ du
e (1 + u2 ) p e u2p
 S E C T I O N 10.3 Convergence of Series with Positive Terms (LT SECTION 11.3) 75

which converges since 2p > 1 (see Theorem 3). If on the other hand p ≤ 1
2, then 1 + u2 ≤ 2u2 , so that
 p
(1 + u2 ) p ≤ 2u2 = 2 p u2p , so that
∞ ∞
1 1 1
du ≥ p du
e (1 + u )
2 p 2 e u2p
which diverges since 2p ≤ 1 (see Theorem 3 in this section, or Theorem 1 in Section 7.7).
Approximating Infinite Sums In Exercises 81–83, let an = f (n), where f is a continuous, decreasing
∞
function such that f (x) ≥ 0 and f (x) dx converges.
1

81. Show that

∞ 
∞ ∞
f (x) dx ≤ an ≤ a1 + f (x) dx 3
1 n=1 1

solution From the proof of the Integral Test, we know that

N ∞
a2 + a3 + a4 + · · · + aN ≤ f (x) dx ≤ f (x) dx
1 1

that is,
∞ ∞
S N − a1 ≤ f (x) dx or S N ≤ a1 + f (x) dx
1 1

Also from the proof of the Integral test, we know that

N
f (x) dx ≤ a1 + a2 + a3 + · · · + aN−1 = S N − aN ≤ S N
1

Thus,
N ∞
f (x) dx ≤ S N ≤ a1 + f (x) dx
1 1

Taking the limit as N → ∞ yields Eq. (3), as desired.

82. Using the inequality in (3), show that
∞
1
5≤ ≤6
n=1
n1.2

This series converges slowly. Use a computer algebra system to verify that S N < 5 for N ≤ 43,128 and
S 43,129 ≈ 5.00000021.
solution By Eq. (3), we have
∞
dx  1 ∞ ∞
dx
1.2
≤ ≤1+
1 x n=1
n1.2 1 x1.2

Since
∞  
dx R
dx 1 R−0.2
= lim = lim − =5
1 x1.2 R→∞ 1 x1.2 R→∞ 0.2 0.2
it follows that
∞
1
5≤ ≤6
n=1
n1.2

Because an = n−1.2 ≥ 0 for all N, S N is increasing and it suffices to show that S N < 5 for N = 43128 to
conclude that S N < 5 for all N ≤ 43128. Using a computer algebra system, we obtain:

43128
1
S 43128 = = 4.9999974685
n=1
n1.2
76 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

and

43129
1
S 43129 = = 5.0000002118
n=1
n1.2


∞
83. Assume an converges to S . Arguing as in Exercise 81, show that
n=1


M ∞ 
M+1 ∞
an + f (x) dx ≤ S ≤ an + f (x) dx 4
n=1 M+1 n=1 M+1

Conclude that
⎛M ⎞
⎜⎜ ∞ ⎟⎟
0 ≤ S − ⎜⎜⎜⎝ an + f (x) dx⎟⎟⎟⎠ ≤ a M+1 5
n=1 M+1

This provides a method for approximating S with an error of at most a M+1 .

solution Following the proof of the Integral Test and the argument in Exercise 81, but starting with n =
M + 1 rather than n = 1, we obtain
∞ 
∞ ∞
f (x) dx ≤ an ≤ a M+1 + f (x) dx
M+1 n=M+1 M+1


M
Adding an to each part of this inequality yields
n=1


M ∞ 
∞ 
M+1 ∞
an + f (x) dx ≤ an = S ≤ an + f (x) dx
n=1 M+1 n=1 n=1 M+1


M ∞
Subtracting an + f (x) dx from each part of this last inequality then gives us
n=1 M+1

⎛M ⎞
⎜⎜ ∞ ⎟⎟
0 ≤ S − ⎜⎜⎜⎝ an + f (x) dx⎟⎟⎟⎠ ≤ a M+1
n=1 M+1

84. Use the inequalities in (4) from Exercise 83 with M = 43,129 to prove that
∞
1
5.5915810 ≤ 1.2
≤ 5.5915839
n=1
n

1 1
solution Using the inequalities in (4) with f (x) = , an = 1.2 and M = 43129, we find
x1.2 n
∞
dx ∞
1 ∞
dx
S 43129 + 1.2
≤ 1.2
≤ S 43130 + 1.2
43130 x n=1
n 43130 x

Now,
S 43129 = 5.0000002118;
1
S 43130 = S 43129 + = 5.0000029551
431301.2
and
∞ R  
dx dx 1 1 5
= lim = −5 lim 0.2 − = = 0.5915808577
43130 x1.2 R→∞ 43130 x 1.2 R→∞ R 431300.2 431300.2
Thus,
∞
1
5.0000002118 + 0.5915808577 ≤ 1.2
≤ 5.0000029551 + 0.5915808577
n=1
n
 S E C T I O N 10.3 Convergence of Series with Positive Terms (LT SECTION 11.3) 77

or
∞
1
5.5915810695 ≤ ≤ 5.5915838128
n=1
n1.2

85. Use the inequalities in (4) from Exercise 83 with M = 40,000 to show that
∞
1
1.644934066 ≤ 2
≤ 1.644934068
n=1
n

Is this consistent with Euler’s result, according to which this infinite series has sum π2 /6?
1 1
solution From the given inequalities with f (x) = 2 , an = 2 and M = 40000, we find
x n
∞
dx ∞
1 ∞
dx
S 40000 + 2
≤ 2
≤ S 40001 + 2
40001 x n=1
n 40001 x

Now,
S 40000 = 1.6449090672;
1
S 40001 = S 40000 + = 1.6449090678
40001
and
∞ R  
dx dx 1 1 1
= lim = − lim − = = 0.0000249994
40001 x2 R→∞ 40001 x 2 R→∞ R 40001 40001
Thus,
∞
1
1.6449090672 + 0.0000249994 ≤ ≤ 1.6449090678 + 0.0000249994
n=1
n2
or
∞
1
1.6449340665 ≤ 2
≤ 1.6449340672
n=1
n

π2
Since ≈ 1.6449340668, our approximation is consistent with Euler’s result.
6

∞
86. Use a CAS and the inequalities in (5) from Exercise 83 to determine the value of n−6 to
n=1
within an error less than 10−4 . Check that your result is consistent with that of Euler, who proved that the
sum is equal to π6 /945.
solution According to the inequalities in (5), if we choose M so that (M + 1)−6 < 10−4 , we can then
approximate the sum to within 10−4 . Solving (M + 1)−6 = 10−4 gives M + 1 = 10−2/3 ≈ 4.641, so the
smallest such integral M is M = 4. Denote by S the sum of the series. Then
⎛ 4 ⎞
⎜⎜⎜ ∞ ⎟⎟⎟
0 ≤ S − ⎜⎜⎜⎝ n−6 + x−6 dx⎟⎟⎟⎠ ≤ (M + 1)−6 < 10−4
n=1 5

We have

4
1 1 1 1
n−6 = + + + ≈ 1.017240883
n=1
1 64 729 4096

∞
−6 1 −5 ∞ 1
x dx = − x  = 6 ≈ 0.000064
5 5 5 5
π6
The sum of these two is ≈ 1.017304883, while ≈ 1.017343063. These two values differ by approxi-
945
−4
mately 0.000038180 < 10 , so the result is consistent with Euler’s calculation.
78 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)


∞
87. Use a CAS and the inequalities in (5) from Exercise 83 to determine the value of n−5 to
n=1
within an error less than 10−4 .

solution Using the inequalities in (5) with f (x) = x−5 and an = n−5 , we have
⎛ M+1 ⎞

∞ ⎜⎜⎜  ∞ ⎟⎟⎟
0≤ n −5
− ⎜⎜⎜⎝ n−5 + x −5
dx⎟⎟⎟⎠ ≤ (M + 1)−5
n=1 n=1 M+1

To guarantee an error less than 10−4 , we need (M + 1)−5 ≤ 10−4 . This yields M ≥ 104/5 − 1 ≈ 5.3, so we
choose M = 6. Now,


7
n−5 = 1.0368498887
n=1

and
∞ R
1  1
x−5 dx = lim x−5 dx = − lim R−4 − 7−4 = = 0.0001041233
7 R→∞ 7 4 R→∞ 4 · 74

Thus,


∞ 
7 ∞
n−5 ≈ n−5 + x−5 dx = 1.0368498887 + 0.0001041233 = 1.0369540120
n=1 n=1 7

88. How far can a stack of identical books (of mass m and unit length) extend without tipping over? The
stack will not tip over if the (n + 1)st book is placed at the bottom of the stack with its right edge located at
or before the center of mass of the first n books (Figure 6). Let cn be the center of mass of the first n books,
measured along the x-axis, where we take the positive x-axis to the left of the origin as in Figure 7. Recall
that if an object of mass m1 has center of mass at x1 and a second object of m2 has center of mass x2 , then
the center of mass of the system has x-coordinate

m1 x1 + m2 x2
m1 + m2

(a) Show that if the (n + 1)st book is placed with its right edge at cn , then its center of mass is located at
cn + 12 .
(b) Consider the first n books as a single object of mass nm with center of mass at cn and the (n + 1)st
book as a second object of mass m. Show that if the (n + 1)st book is placed with its right edge at cn , then
1
cn+1 = cn + .
2(n + 1)
(c) Prove that lim cn = ∞. Thus, by using enough books, the stack can be extended as far as desired without
n→∞
tipping over.

25
≈ 1.04 book lengths
24

1
1 2
1 4
1 6
8 1 1 1 1 25
+ + + =
2 4 6 8 24

FIGURE 6
 S E C T I O N 10.3 Convergence of Series with Positive Terms (LT SECTION 11.3) 79

1
2
3
4

...
1
1
n 1 2
4
n+1 6
1
2(n + 1)
x
cn+1 cn c3 c2 c1 0
FIGURE 7

solution Let f (x) = 1x . This function is continuous, positive and decreasing, so following the argument of
Exercise 81, we know that
N N
f (x) dx ≤ S N ≤ a1 + f (x) dx
1 1
or
1 1 1
ln N ≤ 1 + + + · · · + ≤ 1 + ln N
2 3 N
Using the inequality on the right-hand side, we know that
S 8100 ≤ 1 + ln 8100 = 9.999619 < 10
using the inequality on the left-hand side, we can guarantee S N ≥ 100 by making ln N ≥ 100. Thus, we can
take
N ≥ e100 ≈ 2.688 × 1043

∞
89. The following argument proves the divergence of the harmonic series 1/n without using the Integral
n=1
Test. To begin, assume that the harmonic series converges to a value S .
(a) Prove that the following two series must also converge:
1 1 1 1 1
1+ + + ··· and + + + ···
3 5 2 4 6
(b) Prove that if S 1 and S 2 are the sums of the series on the left and right, respectively, then S = S 1 + S 2 .
(c) Prove that S 1 ≥ S 2 + 12 , and S 2 = 12 S . Explain how this leads to a contradiction and the conclusion that
the harmonic series diverges.
solution
(a) For the first series, let a2k+1 = 2k+1
1
for k ≥ 0 and a2k = 0 for k ≥ 1. Then ak ≤ 1k , so by the Direct
# #
Comparison Test, if the harmonic series converges, so does ak ; but clearly ak is the same as the sum of
the series on the left. For the series on the right, let b2k = 2k1 for k ≥ 1 and b2k+1 = 0 for k ≥ 0. Then bk ≤ 1k ,
#
so again by the Direct Comparison Test convergence of the harmonic series implies convergence of bk , and
therefore of the series on the right.
# # # #
(b) Since ak = S 1 as well, and bk = S 2 , Theorem 2 in Section 10.1 shows that (ak + bk ) = 1k =
S 1 + S 2 . Thus S = S 1 + S 2 .
(c) Note that
1 1 1 1 1 1 1
S2 + = + + + + ··· = 1 + + + ···
2 2 2 4 6 4 6
Then each term of the underlying sequence is less than each term of the underlying sequence for S 1 , so that
S 1 ≥ S 2 + 12 . However,
 
1 1 1 1 1 1 1
S2 = + + + ··· = 1 + + + ··· = S
2 4 6 2 2 3 2
so that S 2 = 12 S . Since S = S 1 + S 2 , this implies that S 1 = 12 S = S 2 . This contradicts the fact that
S 1 ≥ S 2 + 12 , so the original assumption that the harmonic series converges must have been false.
80 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

Further Insights and Challenges


∞
90. Consider the series an , where an = (ln(ln n))− ln n .
n=2
(a) Show, by taking logarithms, that an = n− ln(ln(ln n)) .
e2
(b) Show that ln(ln(ln n)) ≥ 2 if n > C, where C = ee .
(c) Show that the series converges.
solution
(a) Let an = (ln(ln n))− ln n . Then

ln an = (− ln n) ln(ln(ln n)),

and
 − ln(ln(ln n))
an = e(− ln n) ln(ln(ln n)) = eln n = n− ln(ln(ln n)) .

e2
(b) Suppose n > ee . Then
e2 2
ln n > ln ee = ee ;
2
ln(ln n) > ln ee = e2 ; and
ln(ln(ln n)) > ln e2 = 2

(c) Combining the results from parts (a) and (b), we have
1 1
an = ≤
nln(ln(ln n)) n2

e2
∞
1 ∞
1
for n > C = ee . The series 2
is a convergent p-series, so also converges. By the Direct
n=1
n n=C+1
n2
∞
Comparison Test we can therefore conclude that the series an converges, which means that the series
n=C+1

∞
an converges.
n=2


∞
91. Kummer’s Acceleration Method Suppose we wish to approximate S = 1/n2 . There is a similar
n=1
telescoping series whose value can be computed exactly (Example 2 in Section 10.2):


∞
1
=1
n=1
n(n + 1)

(a) Verify that


∞ ∞ 
 
1 1 1
S = + −
n=1
n(n + 1) n=1 n2 n(n + 1)

Thus for M large,


M
1
S ≈1+ 6
n=1
n2 (n + 1)


M
(b) Explain what has been gained. Why is (6) a better approximation to S than 1/n2 ?
n=1
 S E C T I O N 10.3 Convergence of Series with Positive Terms (LT SECTION 11.3) 81

(c) Compute

1000
1 
100
1
, 1+
n=1
n2 n=1
n2 (n+ 1)

Which is a better approximation to S , whose exact value is π2 /6?

solution
∞
1 ∞
1
(a) Because the series and both converge,
n=1
n2
n=1
n(n + 1)
 ∞ 
    1  
∞ ∞ ∞ ∞ ∞
1 1 1 1 1 1
+ − = + − = =S
n=1
n(n + 1) n=1 n2 n(n + 1) n=1
n(n + 1) n=1
n2
n=1
n(n + 1) n=1
n2

Now,
1 1 n+1 n 1
− = 2 − 2 = 2
n2 n(n + 1) n (n + 1) n (n + 1) n (n + 1)
so, for M large,

M
1
S ≈1+
n=1
n2 (n + 1)

#∞ ∞
1
1
(b) The series n=1 n2 (n+1)
converges more rapidly than since the degree of n in the denominator is
n=1
n2
larger.
(c) Using a computer algebra system, we find

1000
1 
100
1
= 1.6439345667 and 1+ = 1.6448848903
n=1
n2 n=1
n2 (n + 1)

π2
The second sum is more accurate because it is closer to the exact solution ≈ 1.6449340668.
6

∞
92. The sum S = k−3 has been computed to more than 100 million digits. The first 30 digits are
k=1

S = 1.202056903159594285399738161511
Approximate S using Kummer’s Acceleration Method of Exercise 91 with the similar series

∞
(n(n + 1)(n + 2))−1
n=1

and M = 500. According to Exercise 60 in Section 10.2, the similar series is a telescoping series with a sum
of 14 .
solution We compute the difference between the general term of the given series and the general term of
the auxiliary series:
1 1 (k + 1)(k + 2) − k2 k2 + 3k + 2 − k2 3k + 2
− = = 3 = 3
k3 k(k + 1)(k + 2) k3 (k + 1)(k + 2) k (k + 1)(k + 2) k (k + 1)(k + 2)
Hence,
∞
1 
∞
1 ∞
3k + 2 1 
∞
3k + 2
= + = +
k=1
k 3
k=1
k(k + 1)(k + 2) k=1
k 3 (k + 1)(k + 2) 4 k=1
k 3 (k + 1)(k + 2)

With M = 100 and using a computer algebra system, we find

∞
1 1 
100
3k + 2
3
≈ + 3 (k + 1)(k + 2)
= 1.2020559349
k=1
k 4 k=1
k
82 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

10.4 Absolute and Conditional Convergence (LT Section 11.4)

Preliminary Questions  
1. Give an example of a series such that an converges but |an | diverges.
 n  1
(−1)
solution The series √3
n
converges by the Alternating Series Test, but the positive series √3 is a
n
divergent p-series.
2. Which of the following statements is equivalent to Theorem 1?

∞ 
∞
(a) If |an | diverges, then an also diverges.
n=0 n=0

∞ 
∞
(b) If an diverges, then |an | also diverges.
n=0 n=0
∞ ∞
(c) If an converges, then |an | also converges.
n=0 n=0

∞ 
∞
solution The correct answer is (b): If an diverges, then |an | also diverges. Take an = (−1)n n1 to see
n=0 n=0
that statements (a) and (c) are not true in general.

∞
√
3. Indicate whether or not the reasoning in the following statement is correct: Since (−1)n n is an alter-
n=1
nating series, it must converge.
∞
√ √
solution No. Although (−1)n n is an alternating series, the terms an = n do not form a decreasing
n=1
√ 
∞
√
sequence that tends to zero. In fact, an = n is an increasing sequence that tends to ∞, so (−1)n n
n=1
diverges by the Divergence Test.

∞
4. Suppose that bn is positive, decreasing, and tends to 0, and let S = (−1)n−1 bn . What can we say about
n=1
|S − S 100 | if a101 = 10−3 ? Is S larger or smaller than S 100 ?
solution From the text, we know that |S − S 100 | < a101 = 10−3 . Also, the Leibniz test tells us that S 2N <
S < S 2N+1 for any N ≥ 1, so that S 100 < S .

Exercises
1. Show that

∞
(−1)n
n=0
2n

converges absolutely.
∞
1 1
solution The positive series n
is a geometric series with r = . Thus, the positive series converges,
n=0
2 2
and the given series converges absolutely.
2. Show that the following series converges conditionally:
∞
1 1 1 1 1
(−1)n−1 2/3 = 2/3 − 2/3 + 2/3 − 2/3 + · · ·
n=1
n 1 2 3 4

solution Let an = n2/3 1

. Then an forms a decreasing sequence that tends to zero; hence, the series
∞
1 ∞
1
(−1)n−1 2/3 converges by the Alternating Series Test. However, the positive series 2/3
is a divergent
n=1
n n=1
n
p-series, so the original series converges conditionally.
 S E C T I O N 10.4 Absolute and Conditional Convergence (LT SECTION 11.4) 83

In Exercises 3–10, determine whether the series converges absolutely, conditionally, or not at all.

∞
(−1)n−1
3.
n=1
n1/3

∞
(−1)n−1
solution The sequence an = 1
n1/3
is positive, decreasing, and tends to zero; hence, the series
n=1
n1/3

∞
1
converges by the Alternating Series Test. However, the positive series is a divergent p-series, so the
n=1
n1/3
original series converges conditionally.
∞
(−1)n n4
4.
n=1
n3 + 1
solution Because

n4
lim =∞
n→∞ n3 + 1

(−1)n n4
the general term 3 of the series does not tend to zero; hence, this series diverges by the Divergence
n +1
Test.
 ∞
(−1)n
5.
n=0
(1.001)n
solution Because
 n 
 (−1)  = 1
 (1.001)n  1.001n

the positive series forms a geometric series with ratio 1

1.001 < 1; thus, the original series converges absolutely.
∞
(−1)n
6.
n=0
(0.999)n
solution Because
1
lim =∞
n→∞ 0.999n
(−1)n
the general term of the series does not tend to zero; hence, this series diverges by the Divergence
0.999n
Test.
 ∞
(−1)n e−n
7.
n=1
n2
solution We have
 n −n 
 (−1) e  = e < 1
−n
 n2  n2 n2
∞
1
since e−n < 1 for n ≥ 1. Therefore the positive series is bounded above by the convergent p-series 2
, so
n=1
n
it is convergent and the original series converges absolutely.
∞
sin( πn
4 )
8. 2
n=1
n
solution Because
     
 sin πn  sin πn 
 4  = 4 1
≤
 n2  n2 n2

the positive series forms a convergent p-series; thus, the original series converges absolutely.
84 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)


∞
(−1)n
9.
n=2
n ln n

solution Let an = 1
n ln n .
Then an forms a decreasing sequence (note that n and ln n are both increasing
 ∞
(−1)n
functions of n) that tends to zero; hence, the series converges by the Alternating Series Test.
n=2
n ln n
∞
1
However, the positive series diverges, so the original series converges conditionally.
n=2
n ln n


∞
(−1)n
10.
n=1 1+ 1
n

solution Because

1 n+1
lim = lim =1
n→∞ 1+ 1
n
n→∞ n

(−1)n
the general term of the series does not tend to zero; hence, this series diverges by the Divergence Test.
1 + 1n

∞
1
11. Let S = (−1)n+1 .
n=1
n3
(a) Calculate S n for 1 ≤ n ≤ 10.
(b) Use the inequality in (2) to show that 0.9 ≤ S ≤ 0.902.
solution
(a)
1
S1 = 1 S6 = S5 − = 0.899782407
63
1 7 1
S2 = 1 − = = 0.875 S7 = S6 + = 0.902697859
23 8 73
1 1
S 3 = S 2 + 3 = 0.912037037 S8 = S7 − = 0.900744734
3 83
1 1
S 4 = S 3 − 3 = 0.896412037 S9 = S8 + = 0.902116476
4 93
1 1
S 5 = S 4 + 3 = 0.904412037 S 10 = S 9 − = 0.901116476
5 103
(b) By (2),

1
|S 10 − S | ≤ a11 =
113
so
1 1
S 10 − ≤ S ≤ S 10 + 3
113 11
or

0.900365161 ≤ S ≤ 0.901867791

12. Use the inequality in (2) to approximate


∞
(−1)n+1
n=1
n!

to four decimal places.

 S E C T I O N 10.4 Absolute and Conditional Convergence (LT SECTION 11.4) 85


∞
(−1)n+1 1
solution Let S = , so that an = . By (2),
n=1
n! n!

1
|S N − S | ≤ aN+1 =
(N + 1)!

To guarantee accuracy to four decimal places, we must choose N so that

1
< 5 × 10−5 or (N + 1)! > 20000
(N + 1)!

Because 7! = 5040 and 8! = 40, 320, the smallest value that satisfies the required inequality is N = 7. Thus,

1 1 1 1 1 1
S ≈ S7 = 1 − + − + − + = 0.632142857
2! 3! 4! 5! 6! 7!


∞
(−1)n+1
13. Approximate to three decimal places.
n=1
n4


∞
(−1)n+1 1
solution Let S = , so that an = . By Eq. (2),
n=1
n4 n4

1
|S N − S | ≤ aN+1 =
(N + 1)4

To guarantee accuracy to three decimal places, we must choose N so that

1 √4
< 5 × 10−4 or N> 2000 − 1 ≈ 5.7
(N + 1)4

The smallest value that satisfies the required inequality is then N = 6. Thus,

1 1 1 1 1
S ≈ S6 = 1 − 4
+ 4 − 4 + 4 − 4 = 0.946767824
2 3 4 5 6

14. Let


∞
n
S = (−1)n−1
n=1
n2 + 1

Use a computer algebra system to calculate and plot the partial sums S n for 1 ≤ n ≤ 100. Observe that the
partial sums zigzag above and below the limit.

n 
∞
solution The partial sums associated with the alternating series (−1)n−1
are plotted below. As
n=1
+1 n2
expected, the partial sums alternate between overestimating and underestimating the sum.

Sn
0.5

0.4

0.3

0.2

0.1 n
0 20 40 60 80 100
86 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

In Exercises 15–16, find a value of N such that S N approximates the series with an error of at most 10−5 .
Using technology, compute this value of S N .

∞
(−1)n+1
15.
n=1
n(n + 2)(n + 3)

∞
(−1)n+1 1
solution Let S = , so that an = . By Eq. (2),
n=1
n (n + 2) (n + 3) n (n + 2) (n + 3)
1
|S N − S | ≤ aN+1 =
(N + 1)(N + 3)(N + 4)
We must choose N so that
1
≤ 10−5 or (N + 1)(N + 3)(N + 4) ≥ 105
(N + 1)(N + 3)(N + 4)
For N = 43, the product on the left-hand side is 95,128, while for N = 44 the product is 101,520; hence, the
smallest value of N which satisfies the required inequality is N = 44. Thus,

44
(−1)n+1
S ≈ S 44 = = 0.0656746
n=1
n(n + 2)(n + 3)


∞
(−1)n+1 ln n
16.
n=1
n!

∞
(−1)n+1 ln n ln n
solution Let S = , so that an = . By Eq. (2),
n=1
n! n!
ln(N + 1)
|S N − S | ≤ aN+1 =
(N + 1)!
To make the error at most 10−5 , we must choose N so that
ln(N + 1)
≤ 10−5
(N + 1)!
For N = 7, the left-hand side of the above inequality is 5.157 × 10−5 , while for N = 8, the left-hand side is
6.055 × 10−6 ; hence, the smallest value for N which satisfies the required inequality is N = 8. Thus,

8
(−1)n+1 ln n
S ≈ S8 = = −0.209975859
n=1
n!

In Exercises 17–32, determine convergence or divergence by any method.


∞
17. 7−n
n=0
1
solution This is a (positive) geometric series with r = < 1, so it converges.
7
∞
1
18. 7.5
n=1
n
solution This is a p-series with p = 7.5 > 1, so it converges.
∞
1
19.
n=1
5 − 3n
n

1
solution Use the Limit Comparison Test with :
5n
1/(5n − 3n ) 5n 1
L = lim = lim = lim =1
n→∞ 1/5 n n→∞ 5 −3
n n n→∞ 1 − (3/5)n
# 1
But ∞ n=1 n is a convergent geometric series. Since L = 1, the Limit Comparison Test tells us that the
5
original series converges as well.
 S E C T I O N 10.4 Absolute and Conditional Convergence (LT SECTION 11.4) 87


∞
1
20.
n=1 n+ 1
n

solution We have

1 n
=
n+ 1
n
n2 + 1

n n 1
Since ≈ = for large n, use the limit comparison test with 1n :
n2 + 1 n2 n

n/(n2 + 1) n2
lim = lim 2 =1
n→∞ 1/n n→∞ n + 1


∞
1
Since the limit exists and is nonzero, and since diverges, the original series diverges as well.
n=1
n

∞
1
21.
n=1
3n4 + 12n
1
solution Use the Limit Comparison Test with :
3n4

(1/(3n4 + 12n) 3n4 1

L = lim = lim = lim =1
n→∞ 1/3n 4 n→∞ 3n + 12n
4 n→∞ 1 + 4n− 3

# 1 #
But ∞ n=1 4
= 13 ∞n=1 n4 is a convergent p-series. Since L = 1, the Limit Comparison Test tells us that the
1
3n
original series converges as well.
∞
(−1)n
22. √
n=1 n2 + 1
1
solution This is an alternating series with an = √ . Because an is a decreasing sequence that con-
n +1
2
 ∞
(−1)n
verges to zero, the series √ converges by the Alternating Series Test.
n=1 n2 + 1
∞
1
23. √
n=1 n +1
2

solution Apply the Limit Comparison Test and compare the series with the divergent harmonic series:

√1
n2 +1 n
L = lim 1
= lim √ =1
n→∞
n
n→∞ n +1
2


∞
1
Because L > 0, we conclude that the series √ diverges.
n=1 n2 + 1
∞
(−1)n n
24. √
n=0 n2 + 1
solution This series diverges, since the general term of the associated positive series tends to 1, not to 0:

n n2 1
lim √ = lim = lim =1
n→∞ n + 1 n→∞
2 n2 + 1 n→∞ 1 + n−2
88 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)


∞
3n + (−2)n
25.
n=1
5n

∞ n ∞  n
 
∞ ∞  n
3 3 (−2)n 2
solution The two series = and = − are both convergent geometric series.
n=1
5n n=1
5 n=1
5n n=1
5
Hence,

∞ ∞  n
 ∞  n
3n + (−2)n 3 2
= + −
n=1
5n n=1
5 n=1
5

also converges.
∞
(−1)n+1
26.
n=1
(2n + 1)!
1
solution This is an alternating series with an = . Because an is a decreasing sequence which
(2n + 1)!
∞
(−1)n+1
converges to zero, the series converges by the Alternating Series Test.
n=1
(2n + 1)!

∞
(−1)n n2 e−n /3
3
27.
n=1

∞
n2 e−n /3
3
solution Consider the associated positive series . This series can be seen to converge by the
n=1
Integral Test:
∞ R 
/3 R
x2 e−x /3
x2 e−x /3
dx = − lim e−x = e−1/3 + lim e−R /3
= e−1/3
3 3 3 3
dx = lim 1
1 R→∞ 1 R→∞ R→∞

The integral converges, so the original series converges absolutely.

∞
ne−n /3
3
28.
n=1

solution This is a positive series, and by the Comparison Test with the associated positive series in the
previous exercise,

∞ 
∞
ne−n /3
n2 e−n /3
3 3
≤
n=1 n=1

Since the series on the right converges, so does the original series.
∞
(−1)n
29.
n=2
n (ln n)2
1/2

1
solution This is an alternating series with an = . Because an is a decreasing sequence which
n1/2 (ln n)2
(−1)n 
∞
converges to zero, the series converges by the Leibniz Test. (Note that the series converges
n=2
n1/2 (ln n)2
1
only conditionally, not absolutely; the associated positive series is eventually greater than 3/4 , which is a
n
divergent p-series.)
 ∞
1
30.
n=2
n(ln n)1/4
solution Use the Integral Test, with the substitution u = ln x:
∞
1 R
1 R
4 3/4 R
1/4
dx = lim 1/4
dx = lim u−1/4 du = lim u ln 2
2 x ln x R→∞ 2 x ln x R→∞ ln 2 R→∞ 3

4 
=− (ln 2)3/4 + lim R3/4
3 R→∞

The integral diverges, so the original series diverges as well.

 S E C T I O N 10.4 Absolute and Conditional Convergence (LT SECTION 11.4) 89

∞
ln n
31.
n=1
n1.05
solution Choose N so that for n ≥ N we have ln n ≤ n0.01 . Then
∞
ln n  n0.01  1
∞ ∞
≤ =
n=N
n1.05 n=N n1.05 n=N n1.04

This is a convergent p-series, so by the Comparison Test, the original series converges as well.
∞
1
32.
n=2
(ln n)2
solution Choose N so that for n ≥ N we have ln n < n0.25 so that ln2 n < n0.5 . Then

∞
1  1
∞
>
n=N
(ln n)2 n=N n0.5

This is a divergent p-series, so by the Comparison Test, the original series diverges as well.
33. Show that
1 1 1 1 1 1
− + − + − + ···
2 2 3 3 4 4
converges by computing the partial sums. Does it converge absolutely?
solution The sequence of partial sums is
1
S1 =
2
1
S2 = S1 − =0
2
1 1
S3 = S2 + =
3 3
1
S4 = S3 − = 0
3
and, in general,
⎧
⎪
⎪
⎪ 1
⎪
⎨N, for odd N
SN =⎪
⎪
⎪
⎪
⎩0, for even N

Thus, lim S N = 0, and the series converges to 0. The positive series is

N→∞

1 1 1 1 1 1 ∞
1
+ + + + + + ··· = 2
2 2 3 3 4 4 n=2
n

which diverges. Therefore, the original series converges conditionally, not absolutely.
34. The Alternating Series Test cannot be applied to
1 1 1 1 1 1
− + − + − + ···
2 3 22 32 23 33
Why not? Show that it converges by another method.
solution The sequence of terms {an } for this alternating series is
1 1 1 1 1 1 1 1 1 1
, , 2 , 2 , 3 , 3 , . . . , n , n , n+1 , n+1 , . . .
2 3 2 3 2 3 2 3 2 3
Now,
1 1 1 1
2
= < = 3
3 9 8 2
90 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

Moreover, if we assume that

1 1
<
3k 2k+1
for some k, then
1 1 1 1 1 1 1 1
= · k < < = k+2
3k+1 3 3 32k+1 22k+1 2
Thus, by mathematical induction,
1 1
<
3n 2n+1
for all n ≥ 2. The sequence {an } is therefore not decreasing, and the Alternating Series Test does not apply.
We may express the given series as
 ∞  
1 1
−
n=1
2n 3n

Because
∞ ∞  n ∞ ∞  n
1 1 1 1
n
= and n
=
n=1
2 n=1
2 n=1
3 n=1
3

are both convergent geometric series, it follows that this series converges, and
∞    ∞  n  ∞  n 1 1
1 1 1 1 3 1 1
− = − = 2
− =1− =
n=1
2n 3n
n=1
2 n=1
3 1− 2 1− 3
1 1 2 2

35. Assumptions Matter Show that the following series diverges:

 
1 1 1 1 1 1 1 1
− + − + − + ··· + − n + ···
2 4 3 8 4 16 n 2
(Note: This demonstrates that in the Alternating Series Test, we need the assumption that the sequence an is
decreasing. It is not enough to assume only that an tends to zero.)
solution Let
 
1 1 1 1 1 1 1 1
R= − + − + − + ··· + − n+1 + · · ·
2 4 3 8 4 16 n+1 2
This is an alternating series with
⎧ 1
⎪
⎪
⎪ , n = 2k − 1
⎪
⎪
⎨k + 1
an = ⎪
⎪
⎪
⎪
⎪ 1
⎩ , n = 2k
2k+1
Note that an → 0 as n → ∞, but the sequence {an } is not decreasing. We will now establish that R diverges.
For sake of contradiction, suppose that R converges. The geometric series
∞
1
2n+1
n=1

converges, so the sum of R and this geometric series must also converge; however,
∞
1 ∞
1
R+ n+1
=
n=1
2 n=2
n

which diverges because the harmonic series diverges. Thus, the series R must diverge.
36. Determine whether the following series converges conditionally:
1 1 1 1 1 1 1 1 1
1− + − + − + − + − + ···
3 2 5 3 7 4 9 5 11
 S E C T I O N 10.4 Absolute and Conditional Convergence (LT SECTION 11.4) 91

solution Although the signs alternate, the terms an are not decreasing, so we cannot apply the Alternating
Series Test. However, we may express the series as
∞    ∞
1 1 n+1
− =
n=1
n 2n + 1 n=1
n(2n + 1)

Using the Limit Comparison Test and comparing with the harmonic series, we find
n+1
n(2n+1) n+1 1
L= lim = lim =
n→∞ 1
n
n→∞ 2n + 1 2

Because L > 0, we conclude that the series

1 1 1 1 1 1 1 1 1
1− + − + − + − + − + ···
3 2 5 3 7 4 9 5 11
diverges.
  
37. Prove that if an converges absolutely, then a2n also converges. Give an example where an is

2
only conditionally convergent and an diverges.
 
solution Suppose the series an converges absolutely. Because |an | converges, we know that

lim |an | = 0
n→∞

Therefore, there exists a positive integer N such that |an | < 1 for all n ≥ N. It then follows that for n ≥ N,

0 ≤ a2n = |an |2 = |an | · |an | < |an | · 1 = |an |


By the Direct Comparison Test we can then conclude that a2n also converges.
 ∞
(−1)n
Consider the series √ . This series converges by the Alternating Series Test, but the correspond-
n=1
n
∞
(−1)n ∞
ing positive series is a divergent p-series; that is, √ is conditionally convergent. Now, a2n is the
n=1
n n=1
∞
1  
divergent harmonic series . Thus, a2n need not converge if an is only conditionally convergent.
n=1
n

Further Insights and Challenges

38. Prove the following variant of the Alternating Series Test: If {bn } is a positive, decreasing sequence with
lim bn = 0, then the series
n→∞

b1 + b2 − 2b3 + a4 + b5 − 2a6 + · · ·

converges. Hint: Show that S 3N is increasing and bounded by a1 + a2 , and continue as in the proof of the
Alternating Series Test.
solution Following the hint, we first examine the sequence {S 3N } (where {S N } is the given sequence). Now,

S 3N+3 = S 3(N+1) = S 3N + a3N+1 + a3N+2 − 2a3N+3 = S 3N + (a3N+1 − a3N+3 ) + (a3N+2 − a3N+3 ) ≥ S 3N

because {an } is a decreasing sequence. Moreover,


N−1
S 3N = a1 + a2 − (2a3k − a3k+1 − a3k+2 ) − 2a3N
k=1


N−1
= a1 + a2 − [(a3k − a3k+1 ) + (a3k − a3k+2 ) − 2a3N ] ≤ a1 + a2
k=1

again because {an } is a decreasing sequence. Thus, {S 3N } is an increasing sequence with an upper bound;
hence, {S 3N } converges. Next,

S 3N+1 = S 3N + a3N+1 and S 3N+2 = S 3N + a3N+1 + a3N+2

92 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

Given that lim an = 0, it follows that

n→∞

lim S 3N+1 = lim S 3N+2 = lim S 3N

N→∞ N→∞ N→∞

Having just established that lim S 3N exists, it follows that the sequences {S 3N+1 } and {S 3N+2 } converge to
N→∞
the same limit. Finally, we can conclude that the sequence of partial sums {S N } converges, so the given series
converges.
39. Use Exercise 38 to show that the following series converges:
1 1 2 1 1 2
+ − + + − + ···
ln 2 ln 3 ln 4 ln 5 ln 6 ln 7
solution The given series has the structure of the generic series from Exercise 38 with an = ln(n+1)
1
. Because
an is a positive, decreasing sequence with lim an = 0, we can conclude from Exercise 38 that the given series
n→∞
converges.
40. Prove the conditional convergence of
1 1 3 1 1 1 3
1+ + − + + + − + ···
2 3 4 5 6 7 8
solution Using Exercise 38 as a template, we first examine the sequence {R4N } (where {RN } is the given
sequence). Now,
1 1 1 3
R4N+4 = R4(N+1) = R4N + + + −
4N + 1 4N + 2 4N + 3 4N + 4
     
1 1 1 1 1 1
= RN + − + − + − ≥ R4N
4N + 1 4N + 4 4N + 2 4N + 4 4N + 3 4N + 4
Moreover,
N−1  
1 1  3 1 1 1 3 1 1
R4N =1+ + − − − − − ≤1+ +
2 3 k=1 4k 4k + 1 4k + 2 4k + 3 4N 2 3

Thus, {R4N } is an increasing sequence with an upper bound; hence, {R4N } converges. Next,
1
R4N+1 = R4N + ;
4N + 1
1 1
R4N+2 = R4N + + ; and
4N + 1 4N + 2
1 1 1
R4N+3 = R4N + + +
4N + 1 4N + 2 4N + 3
so

lim R4N+1 = lim R4N+2 = lim R4N+3 = lim R4N

n→∞ N→∞ N→∞ N→∞

Having just established that lim R4N exists, it follows that the sequences {R4N+1 }, {R4N+2 } and {R4N+3 } con-
N→∞
verge to the same limit. Finally, we can conclude that the sequence of partial sums {RN } converges, so the
series R converges.
Now, consider the positive series
1 1 3 1 1 1 3
R+ = 1 + + + + + + + + ···
2 3 4 5 6 7 8
Because the terms in this series are greater than or equal to the corresponding terms in the divergent har-
monic series, it follows from the Direct Comparison Test that R+ diverges. Thus, by definition, R converges
conditionally.
41. Show that the following series diverges:
1 1 2 1 1 1 2
1+ + − + + + − + ···
2 3 4 5 6 7 8
Hint: Use the result of Exercise 40 to write the series as the sum of a convergent series and a divergent series.
 S E C T I O N 10.4 Absolute and Conditional Convergence (LT SECTION 11.4) 93

solution Let
1 1 3 1 1 1 3
R=1+ + − + + + − + ···
2 3 4 5 6 7 8
and
1 1 2 1 1 1 2
S =1+ + − + + + − + ···
2 3 4 5 6 7 8
For sake of contradiction, suppose the series S converges. From Exercise 40, we know that the series R
converges. Thus, the series S − R must converge; however,

11
∞
1 1 1
S −R= + + + ··· =
4 8 12 4 k=1 k

which diverges because the harmonic series diverges. Thus, the series S must diverge.
42. Prove that
∞
(ln n)a
(−1)n+1
n=1
n

converges for all exponents a. Hint: Show that f (x) = (ln x)a /x is decreasing for x sufficiently large.
(ln n)a (ln x)a
solution This is an alternating series with an = n . Following the hint, consider the function f (x) = .
x
Now,
a(ln x)a−1 − (ln x)a (ln x)a−1
f  (x) = = (a − ln x)
x2 x2
so f  (x) < 0 and f is decreasing for x > ea . If a ≤ 0, then it is clear that
(ln x)a
lim =0
x→∞ x
if a > 0, then repeated use of L’Hôpital’s Rule leads to the same conclusion. Let N be any integer greater than
∞
(ln n)a
e ; then, {an } is a decreasing sequence for n ≥ N which converges to zero and the series
a
(−1)n+1
n=N
n
∞
(ln n) a
converges by the Alternating Series Test. Finally, the series (−1)n+1 also converges.
n=1
n
43. We say that {bn } is a rearrangement of {an } if {bn } has the same terms as {an } but occurring in a different
∞ 
∞
order. Show that if {bn } is a rearrangement of {an } and an converges absolutely, then bn also converges
n=1 n=1

∞
absolutely. (This result does not hold if an is only conditionally convergent.) Hint: Prove that the partial
n=1

N
sums |bn | are bounded. It can be shown further that the two series converge to the same value.
n=1

∞
solution Suppose the series S = an converges absolutely and denote the corresponding positive series
n=1
by

∞
S+ = |an |
n=1


N 
∞
Further, let T N = |bn | denote the Nth partial sum of the series |bn |. Because {bn } is a rearrangement of
n=1 n=1
{an }, we know that

∞
0 ≤ TN ≤ |an | = S +
n=1
94 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

that is, the sequence {T N } is bounded. Moreover,


N+1
T N+1 = |bn | = T N + |bN+1 | ≥ T N
n=1


∞
that is, {T N } is increasing. It follows that {T N } converges, so the series |bn | converges, which means the
n=1

∞
series bn converges absolutely.
n=1
44. Assumptions Matter In 1829, Lejeune Dirichlet pointed out that the great French mathematician Au-
gustin Louis Cauchy made a mistake in a published paper by improperly assuming the Limit Comparison
Test to be valid for nonpositive series. Here are Dirichlet’s two series:
∞ ∞  
(−1)n (−1)n (−1)n
√ , √ 1+ √
n=1
n n=1
n n
Explain how they provide a counterexample to the Limit Comparison Test when the series are not assumed
to be positive.
solution Let
 ∞  ∞  
(−1)n (−1)n (−1)n
R= √ and S = √ 1+ √
n=1
n n=1
n n
R is an alternating series that converges by the Alternating Series Test; however, we cannot apply the Alter-
nating Series Test to S because the absolute value of the terms in S is not decreasing. Because
 
(−1)n
√ 1 + (−1)n
√  
n n (−1)n
L = lim (−1)n
= lim 1 + √ =1
n→∞ √ n→∞ n
n

if the Limit Comparison Test were valid for nonpositive series, we would conclude that S converges. How-
ever, if we assume that S converges, then the series S − R would also converge. But
∞    ∞
(−1)n 1 (−1)n 1
S −R= √ + − √ =
n=1
n n n n=1
n
which is the divergent harmonic series. Thus, S diverges, and the Limit Comparison Test is not valid for
nonpositive series.

10.5 The Ratio and Root Tests and Strategies for Choosing Tests
(LT Section 11.5)
Preliminary Questions

∞
1. Consider the geometric series crn .
n=0  
(a) In the Ratio Test, what do the terms  aan+1  equal?
√n n
(b) In the Root Test, what do the terms |an | equal?
solution
   
 an+1   crn+1 
(a)  =  = |r|
a   crn 
n n n n n n
(b) |an | = |crn | = |c| · |rn | = |r| |c|
 ∞
2. Consider the p-series n−p .
n=1  
(a) In the Ratio Test, what do the terms  aan+1 n
 equal?
(b) What can be concluded from the Ratio Test?
 S E C T I O N 10.5 The Ratio and Root Tests and Strategies for Choosing Tests (LT SECTION 11.5) 95

solution
   
 an+1   (n + 1)−p   n  p
(a)  = =
an   n−p  n+1
(b) Since
 n p  n p
lim = lim =1
n→∞ n + 1 n→∞ n + 1

the Ratio Test is inconclusive.

∞
1 ∞
1
3. Is the Ratio Test conclusive for ? Is it conclusive for ?
n=1
n! n=1
n + 1
∞
1 1
solution The general term of is an = ; thus,
n=1
n! n!
 
 an+1  = 1 n! 1
 an  (n + 1)! · 1 = n + 1

and
 
an+1 
ρ = lim 
1
 = lim =0<1
n→∞ an  n→∞ n + 1

∞
1
Consequently, the Ratio Test guarantees that the series converges.
n=1
n!
∞
1 1
The general term of is an = ; thus,
n=1
n + 1 n + 1
 
 an+1  = 1 · n + 1 = n + 1
 an  n + 2 1 n+2
and
 
 an+1  n+1
ρ = lim   = lim =1
n→∞ an  n→∞ n + 2


∞
1
The Ratio Test is therefore inconclusive for the series .
n=1
n+1

∞ ∞  −n
1 1
4. Is the Root Test conclusive for ? Is it conclusive for 1 + ?
n=1
2n n=1
n

∞
1 1
solution The general term of n
is an = n ; thus,
n=1
2 2
 
 an+1  = 1 · 2 = 1
n
 an  2n+1 1 2
and
 
an+1  1
ρ = lim  = <1
n→∞ an  2
∞
1
Consequently, the Ratio Test guarantees that the series converges.
2n
∞  −n n=1
 1
The general term of 1+ is
n=1
n
 −n  −n  n n
1 n+1
1+ = =
n n n+1
96 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

thus,
   n+1  n
 an+1  = n + 1 ·
n+1
=
(n + 1)2n+1
 an  n+2 n (n + 2)n+1 nn
and
 
an+1  (n + 1)2n+1 (1 + 1/n)2n+1
ρ = lim   = lim = lim =1
n→∞ an n→∞ (n + 2) n+1 nn n→∞ (1 + 2/n)n+1 1n

 ∞  −n
1
The Ratio Test is therefore inconclusive for the series 1+ .
n=1
n

Exercises
In Exercises 1–20, apply the Ratio Test to determine convergence or divergence, or state that the Ratio Test
is inconclusive.
∞
1
1.
n=1
5n

solution With an = 1
5n ,
   
 an+1  = 1 · 5 = 1 an+1  1
n
and ρ = lim 
 = <1
 an  5n+1 1 5 n→∞ an  5
∞
1
Therefore, the series n
converges by the Ratio Test.
n=1
5

∞
(−1)n−1 n
2.
n=1
5n
(−1)n−1 n
solution With an = 5n ,
   
 an+1  = n + 1 · 5 = n + 1 an+1  1
n
 an  and ρ = lim  = <1
5 n+1 n 5n n→∞ an  5

∞
(−1)n−1 n
Therefore, the series converges by the Ratio Test.
n=1
5n
∞
1
3. n
n=1
n
(−1)n−1
solution With an = nn ,
   −n
 an+1  = 1 nn 1  n n 1 1
·
 an  (n + 1)n+1 1 = = 1 +
n+1 n+1 n+1 n
and
 
an+1 
ρ = lim 
1
=0· =0<1
n→∞ an  e

∞
(−1)n−1
Therefore, the series converges by the Ratio Test.
n=1
nn
∞
3n + 2
4. 3+1
n=0
5n

solution With an = 3n+2

5n3 +1
,
 
 an+1  = 3(n + 1) + 2 · 5n + 1 = 3n + 5 · 5n + 1
3 3
 an  5(n + 1)3 + 1 3n + 2 3n + 2 5(n + 1)3 + 1
 S E C T I O N 10.5 The Ratio and Root Tests and Strategies for Choosing Tests (LT SECTION 11.5) 97

and
 
an+1 
ρ = lim  =1·1=1
n→∞ an 

 ∞
3n + 2
Therefore, for the series , the Ratio Test is inconclusive.
n=0
5n3 + 1
We can show that this series converges by using the Limit Comparison Test and comparing with the
∞
1
convergent p-series .
n=1
n2

∞
n
5.
n=1
n2 +1

solution With an = n
n2 +1
,
 
 an+1  = n+1
·
n2 + 1 n + 1
= ·
n2 + 1
 an  (n + 1)2 + 1 n n n2 + 2n + 2
and
 
an+1 
ρ = lim  =1·1=1
n→∞ an 

n 
∞
Therefore, for the series , the Ratio Test is inconclusive.
n=1
+1 n2
We can show that this series diverges by using the Limit Comparison Test and comparing with the diver-
gent harmonic series.

∞
2n
6.
n=1
n
2n
solution With an = n,
   
 an+1  = 2 an+1 
n+1
ρ = lim 
n 2n
 an  n + 1 · 2n = n + 1 and =2>1
n→∞ an 


∞
2n
Therefore, the series diverges by the Ratio Test.
n=1
n
∞
2n
7.
n=1
n100
2n
solution With an = n100
,
   n 100  
 an+1  = 2 an+1 
n+1
n100
·
 an  (n + 1)100 2n = 2 and ρ = lim   = 2 · 1100 = 2 > 1
n+1 n→∞ an 

∞
2n
Therefore, the series diverges by the Ratio Test.
n=1
n100
∞
n3
8.
n=1
3n2
n3
solution With an = ,
3n2
   3  
 an+1  = (n + 1) · 3 = n + 1 · 1 an+1 
3 n2
 an  and ρ = lim   = 13 · 0 = 0 < 1
3(n+1)2 n3 n 32n+1 n→∞ an 

∞
n3
Therefore, the series converges by the Ratio Test.
n=1
3n2
98 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)


∞
10n
9.
n=1
2n2
10n
solution With an = ,
2n2
   
 an+1  = 10 an+1 
2
n+1
2n
ρ = lim 
1
 an  2(n+1)2 10n = 10 · 22n+1
· and  = 10 · 0 = 0 < 1
n→∞ an 


∞
10n
Therefore, the series converges by the Ratio Test.
n=1
2n2

∞
en
10.
n=1
n!
n
solution With an = ne! ,
   
 an+1  = e an+1 
n+1
ρ = lim 
n! e
 an  (n + 1)! · en = n + 1 and =0<1
n→∞ an 


∞
en
Therefore, the series converges by the Ratio Test.
n=1
n!
∞
en
11.
n=1
nn
en
solution With an = nn ,
   −n
 an+1  = en+1 nn e  n n e 1
 an  (n + 1)n+1 · en = n + 1 n + 1 = n + 1 1 + n

and
 
an+1 
ρ = lim 
1
=0· =0<1
n→∞ an  e
∞
en
Therefore, the series converges by the Ratio Test.
n=1
nn

∞
n40
12.
n=1
n!
50
solution With an = nn! ,
   50  50
 an+1  = (n + 1) · n! = 1 n+1
50
1 1
 an  = 1+
(n + 1)! n50 n + 1 n n+1 n
and
 
an+1 
ρ = lim  =0·1=0<1
n→∞ an 


∞
n50
Therefore, the series converges by the Ratio Test.
n=1
n!
∞
n!
13. n
n=0
6

solution With an = (−1)n n4!n ,

   
 an+1  = (n + 1)! · 4 = n + 1 an+1 
n
 an  and ρ = lim  =∞>1
4n+1 n! 4 n→∞ an 


∞
n!
Therefore, the series (−1)n diverges by the Ratio Test.
n=0
4n
 S E C T I O N 10.5 The Ratio and Root Tests and Strategies for Choosing Tests (LT SECTION 11.5) 99

∞
n!
14.
n=1
n9

solution With an = nn!4 ,

   
 an+1  = (n + 1)! · n = an+1 
4
n4 
 an  (n + 1)4 n! (n + 1)3 and ρ = lim  =∞>1
n→∞ an 

∞
n!
Therefore, the series 4
diverges by the Ratio Test.
n=1
n

∞
1
15.
n=2
n3/2 ln n
1
solution With an = ,
n3/2 ln n
 
 an+1  = 1 n3/2 ln n  n 3/2 ln n
 an  (n + 1)3/2 ln(n + 1) · =
1 n+1 ln(n + 1)
and
 
an+1 
ρ = lim 
ln n
 = 1 · lim
n→∞ an n→∞ ln(n + 1)

Now,
ln n ln x 1/x x+1
lim = lim = lim = lim =1
n→∞ ln(n + 1) x→∞ ln(x + 1) x→∞ 1/(x + 1) x→∞ x

∞
1
Thus ρ = 1 and the Ratio Test is inconclusive for the series .
n=2
n3/2 ln n
1 
∞
1
Using the Integral Test, or the Comparison Test against , we can show that the series
n3/2 n=2
n3/2ln n
converges.

∞
1
16.
n=1
(2n)!

solution With an = 1
(2n)!
,
   
 an+1  = 1 (2n)! 1
ρ = lim 
an+1 
=0<1
 an  (2n + 2)! · 1 = (2n + 2)(2n + 1) and
n→∞ an 


∞
1
Therefore, the series converges by the Ratio Test.
n=1
(2n)!

∞
n 2
17.
n=1
(2n + 1)!
n2
solution With an = (2n+1)!
,
   2
 an+1  = (n + 1) · (2n + 1)! = n + 1
2
1
 an  (2n + 3)! n2 n (2n + 3)(2n + 2)
and
 
 an+1 
ρ = lim   = 12 · 0 = 0 < 1
n→∞ an 


∞
n2
Therefore, the series converges by the Ratio Test.
n=1
(2n + 1)!
100 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

∞
(n!)3
18.
n=1
(3n)!
(n!)3
solution With an = ,
(3n)!
 
 an+1  = ((n + 1)!) · (3n)! = (n + 1)3 n3 + 3n2 + 3n + 1
3
 an  (3(n + 1))! (n!)3 =
(3n + 3)(3n + 2)(3n + 1) 27n3 + 54n2 + 33n + 6
1 + 3n−1 + 3n−2 + 1n−3
=
27 + 54n−1 + 33n−2 + 6n−3
and
 
an+1 
ρ = lim 
1
= <1
n→∞ an  27
∞
(n!)3
Therefore, the series converges by the Ratio Test.
n=1
(3n)!

∞
1
19.
n=2
2n + 1
1
solution With an = ,
2n +1
 
 an+1  = 1 2n + 1 1 + 2−n
 an  2n+1 + 1 · 1 = 2 + 2−n
and
 
an+1  1
ρ = lim  = <1
n→∞ an  2

∞
1
Therefore, the series converges by the Ratio Test.
n=2
2n + 1
∞
1
20.
n=2
ln n
1
solution With an = ,
ln n
 
 an+1  = 1 ln n ln n
 an  ln(n + 1) · 1 = ln(n + 1)
Then using L’Hôpital’s Rule,
ln n 1/n n+1
ρ = lim = lim = lim =1
n→∞ ln(n + 1) n→∞ 1/(n + 1) n→∞ n
Since ρ = 1, the Ratio Test is inconclusive.
The Integral Test would show that this series diverges.

∞
21. Show that nk 3−n converges for all exponents k.
n=1
solution With an = nk 3−n ,
   k
 an+1  = (n + 1) 3
k −(n+1)
1 1
 an  = 1+
nk 3−n 3 n
and, for all k,
 
 an+1  1 1
ρ = lim  = ·1= <1
n→∞ an  3 3

∞
Therefore, the series nk 3−n converges for all exponents k by the Ratio Test.
n=1
 S E C T I O N 10.5 The Ratio and Root Tests and Strategies for Choosing Tests (LT SECTION 11.5) 101


∞
22. Show that n2 xn converges if |x| < 1.
n=1

solution With an = n2 xn ,
   2  
 an+1  = (n + 1) |x| an+1 
2 n+1
= 1+
1
|x| and ρ = lim 
  = 1 · |x| = |x|
 an  n2 |x|n n n→∞ an 


∞
Therefore, by the Ratio Test, the series n2 xn converges provided |x| < 1.
n=1


∞
23. Show that 2n xn converges if |x| < 12 .
n=1

solution With an = 2n xn ,
   
 an+1  = 2 |x| an+1 
n+1 n+1
 an  = 2|x| and ρ = lim   = 2|x|
2n |x|n n→∞ an 


∞
Therefore, ρ < 1 and the series 2n xn converges by the Ratio Test provided |x| < 12 .
n=1


∞
rn
24. Show that converges for all r.
n=1
n!
n
solution With an = nr! ,
   
 an+1  = |r| |r| an+1 
n+1
n! 
  = 0 · |r| = 0 < 1
 an  (n + 1)! · |r|n = n + 1 and ρ = lim 
n→∞ an 


∞
rn
Therefore, the series converges by the Ratio Test for all r.
n=1
n!


∞
rn
25. Show that converges if |r| < 1.
n=1
n
rn
solution With an = n,
   
 an+1  = |r| an+1 
n+1
ρ = lim 
n n
 an  n + 1 · |r|n = |r| n + 1 and  = 1 · |r| = |r|
n→∞ an 


∞
rn
Therefore, by the Ratio Test, the series converges provided |r| < 1.
n=1
n


∞
2n
26. Is there any value of k such that converges?
n=1
nk
2n
solution With an = nk
,
   n k
 an+1  = 2
n+1
nk
 an  (n + 1)k 2n· = 2
n+1
and, for all k,
 
 an+1 
ρ = lim   = 2 · 1k = 2 > 1
n→∞ an 


∞
2n
Therefore, by the Ratio Test, there is no value for k such that the series converges.
n=1
nk
102 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)
 n
1
In Exercises 27–28, the following limit could be helpful: lim 1 + = e.
n→∞ n

∞
n!
27. Does n
converge or diverge?
n=1
n

solution With an = nn!n ,

   n n  −n
 an+1  = (n + 1)! · n =
n
1
 an  (n + 1)n+1 n! = 1 +
n+1 n

and
 
 an+1  1
ρ = lim  = <1
n→∞ an  e
∞
n!
Therefore, the series converges by the Ratio Test.
n=1
nn


∞
(2n)!
28. Does converge or diverge?
n=1
nn

(2n)!
solution With an = nn ,

     −n
 an+1  = (2n + 2)(2n + 1)(2n)! · n = 2(2n + 1) n n = 2(2n + 1) 1 + 1
n
 an  (n + 1)n+1 (2n)! n+1 n

and
 
an+1 
ρ = lim  =∞
n→∞ an 

 −n ∞
1 1 (2n)!
since limn→∞ 2(2n + 1) diverges but lim 1 + = . Therefore, the series diverges.
n→∞ n e n=1
nn

In Exercises 29–33, assume that |an+1 /an | converges to ρ = 13 . What can you say about the convergence of
the given series?


∞
29. n3 an
n=1
solution Let bn = n3 an . Then
   3  
 bn+1  n + 1  an+1  1 1
ρ = lim   = lim   = 13 · = < 1
n→∞ bn n→∞ n an 3 3


∞
Therefore, the series n3 an converges by the Ratio Test.
n=1

∞
30. 2n an
n=1
solution Let bn = 2n an . Then
   
bn+1  2n+1  an+1  = 2 · 1 = 2 < 1
ρ = lim   = lim n  an 
n→∞ bn n→∞ 2 3 3

∞
Therefore, the series 2n an converges by the Ratio Test.
n=1
 S E C T I O N 10.5 The Ratio and Root Tests and Strategies for Choosing Tests (LT SECTION 11.5) 103


∞
31. 3n an
n=1

solution Let bn = 3n an . Then

   
bn+1  3n+1  an+1  = 3 · 1 = 1
ρ = lim   = lim n  an 
n→∞ bn n→∞ 3 3

∞
Therefore, the Ratio Test is inconclusive for the series 3n an .
n=1


∞
32. 4n an
n=1

solution Let bn = 4n an . Then

   
bn+1  4n+1  an+1  = 4 · 1 = 4 > 1
ρ = lim   = lim n  an 
n→∞ bn n→∞ 4 3 3

∞
Therefore, the series 4n an diverges by the Ratio Test.
n=1


∞
33. a2n
n=1

solution Let bn = a2n . Then

     2
 bn+1   an+1 2 1 1
ρ = lim   = lim   = = <1
n→∞ bn  n→∞ an  3 9


∞
Therefore, the series a2n converges by the Ratio Test.
n=1


∞
34. Assume that |an+1 /an | converges to ρ = 4. Does a−1
n converge (assume that an  0 for all n)?
n=1

solution Let bn = a−1

n . Then
   
 bn+1   an  1 1
ρ = lim   = lim  =   = < 1
n→∞ bn  n→∞ an+1 
limn→∞  an 
an+1 4


∞
Therefore, the series a−1
n converges by the Ratio Test.
n=1

∞
1
35. Show that the root test is inconclusive for the p-series .
n=1
np

1
solution With an = = n−p ,
np
√n
lim n−p = lim n−p/n
n→∞ n→∞

To compute this limit, we compute the limit of its logarithm and use L’Hôpital’s rule:
 −p ln n −p/n −p
lim ln n−p/n = lim = lim = lim =0
n→∞ n→∞ n n→∞ 1 n→∞ n

Since the limit of the logarithm is zero, the limit of the original expression is 1, so the root test is inconclusive.
104 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

In Exercises 36–41, use the Root Test to determine convergence or divergence (or state that the test is incon-
clusive).
∞
1
36.
n=0
10n
solution With an = 1
10n ,

√n n 1 1 √n 1
an = n
= and lim an = <1
10 10 n→∞ 10
∞
1
Therefore, the series n
converges by the Root Test.
n=0
10
∞
1
37. n
n=1
n
solution With an = 1
nn ,

√n n 1 1 √n
an = = and lim an = 0 < 1
nn n n→∞

∞
1
Therefore, the series n
converges by the Root Test.
n=0
n
∞  k
k
38.
k=0
k + 10
 k
solution With ak = k+10 k
,

 k
√k k k k √k
ak = = and lim ak = 1
k + 10 k + 10 k→∞

∞ 
 k
k
Therefore, the Root Test is inconclusive for the series . Because
k=0
k + 10

 −k ⎡ k/10 ⎤−10

10 ⎢⎢⎢ 10 ⎥⎥⎥
lim ak = lim 1 + ⎢
= lim ⎢⎣ 1 + ⎥⎥⎦ = e−10  0
k→∞ k→∞ k k→∞ k

this series diverges by the Divergence Test.

∞  k
k
39.
k=0
3k + 1
 k
solution With ak = 3k+1 k
,

 k
√k k k k √k 1
ak = = and lim ak = < 1
3k + 1 3k + 1 k→∞ 3
∞ 
 k
k
Therefore, the series converges by the Root Test.
k=0
3k + 1
∞  −n
1
40. 2+
n=1
n
 −n
1
solution With ak = 2 + ,
n
 −n  −1
√n 1 1 n
an = 2+ = 2+ =
n

n n 2n + 1
 S E C T I O N 10.5 The Ratio and Root Tests and Strategies for Choosing Tests (LT SECTION 11.5) 105

Then
√n n 1
lim an = lim = <1
n→∞ n→∞ 2n + 1 2
∞ 
 −n
1
Therefore the series 2+ converges by the Root Test.
n=1
n
∞ 
 −n2
1
41. 1+
n=4
n
1 −n
2
solution With ak = 1 + n ,

 −n2  −n
√n 1 1 √n
an = e−1 < 1
n
an = 1+ = 1+ and lim
n n n→∞

∞ 
 −n2
1
Therefore, the series 1+ converges by the Root Test.
k=0
n


∞
2n
2
2
42. Prove that diverges. Hint: Use 2n = (2n )n and n! ≤ nn .
n=1
n!

solution Because n! ≤ nn ,
2 2
2n 2n
≥ n
n! n
2
2n
Now, let an = nn . Then

√n n 2n2 2n
an = =
nn n
and
√n 2n 2x 2 x ln 2
lim an = lim = lim = lim =∞>1
n→∞ n→∞ n x→∞ x x→∞ 1

∞
2n
2

Therefore, the series diverges by the Root Test. By the Direct Comparison Test, we can then conclude
n=1
nn

∞
2n
2

that the series also diverges.

n=1
n!

In Exercises 43–62, determine convergence or divergence using any method covered in the text so far.


∞
2n + 4n
43.
n=1
7n
solution Because the series


∞ ∞  n
 
∞ ∞  n

2n 2 4n 4
= and =
n=1
7n n=1
7 n=1
7n n=1
7

are both convergent geometric series, it follows that


∞ ∞  n
 ∞  n

2n + 4n 2 4
= +
n=1
7n n=1
7 n=1
7

also converges.
106 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)


∞
n3
44.
n=1
n!
3
solution The presence of the factorial suggests applying the Ratio Test. With an = nn! ,
   
 an+1  = (n + 1) · n! = (n + 1)  an+1  = 0 < 1
3 2
 an  (n + 1)! n3 and ρ = lim
n3 n→∞  an 

 ∞
n3
Therefore, the series converges by the Ratio Test.
n=1
n!
∞
n
45.
n=1
2n + 1
solution Since
n 1 1
lim = lim =
n→∞ 2n + 1 n→∞ 2 + 1/n 2
this series fails the nth Term Divergence Test, so it diverges.
∞
46. 21/n
n=1
solution This series fails the nth Term Divergence Test, so it diverges. To see this, note that 21/n > 1 for
any positive integer n, so that certainly lim 21/n  0.
n→∞

∞
sin n
47.
n=1
n2
  ∞ 1
solution Since |sin x| ≤ 1 for any x, we have  sinn2n  ≤ n12 . Since is a p-series with p > 1, it
n=1 n2
converges. Then by the Direct Comparison Test,
∞  
 sin n 
 n2 
n=1

converges as well, so that the original series converges absolutely.

∞
n!
48.
n=1
(2n)!
solution Use the Ratio Test:
   
 an+1   (n + 1)! (2n)!  n+1
lim   = lim  ·  = lim
n→∞  an  n→∞ (2(n + 1))! n!  n→∞ (2n + 2)(2n + 1)
n+1 1 + 1/n
= lim = lim =0
n→∞ 4n2 + 6n + 2 n→∞ 4n + 6 + 2/n
So by the Ratio Test, the series converges.
∞
1
49. √
n=1
n+ n
√ √
solution Use the Direct Comparison Test. Since n + n ≥ 2 n for n ≥ 1, we have 1√
n+ n
≤ 1
√
2 n
. But
∞
1 1 1
∞
√ =
n=1
2 n 2 n=1 n1/2
∞ 1
diverges, since it is a p-series with p ≤ 1, so that √ diverges as well.
n=1 n + n
 ∞
1
50.
n=2
n(ln n)3
solution By Exercise 79 in Section 10.3, this series converges. To show this directly, use the Integral Test:
   
∞
1 u=ln x
∞
1 1 b 1 1 1
3
dx = 3
du = lim − 2  = lim 2
− 2 =
2 x(ln x) ln 2 u b→∞ 2u ln 2 b→∞ 2(ln 2) 2b 2(ln 2)2
Since the integral exists, the series converges.
 S E C T I O N 10.5 The Ratio and Root Tests and Strategies for Choosing Tests (LT SECTION 11.5) 107


∞
n3
51.
n=1
5n
n3
solution The presence of the exponential term suggests applying the Ratio Test. With an = 5n ,
   3  
 an+1  = (n + 1) · 5 = 1 1 + 1  an+1  = 1 · 13 = 1 < 1
3 n
 an  and ρ = lim
5n+1 n3 5 n n→∞  an  5 5

∞
n3
Therefore, the series converges by the Ratio Test.
n=1
5n

∞
1
52.
n=2
n2 (ln n)3
1
solution The general term in this series suggests applying the Limit Comparison Test. Letting an =
n2 (ln n)3
1
and bn = , we get
n2
an 1/(n2 (ln n)3 ) 1
= lim
lim = lim =0
n→∞ bn n→∞ 1/n2 n→∞ (ln n)3
# #
Since bn converges (it is a p-series with p = 2), an converges as well.
∞
1
53. √
n=2 n − n2
3

solution This series is similar to a p-series; because

1 1 1
√ ≈ √ = 3/2
n3 − n2 n 3 n
for large n, we will apply the Limit Comparison Test comparing with the p-series with p = 32 . Now,
√ 1
n3 −n2 n3
L = lim = lim =1
n→∞ 1
n3/2
n→∞ n3 − n2

∞
1
The p-series with p = 3
2 converges and L exists; therefore, the series √ also converges.
n=2 n3 − n2

∞
n2 + 4n
54.
n=1
3n4 + 9
solution This series is similar to a p-series; because
n2 + 4n n2 1
≈ √ = 2
3n + 9
4
3n4 3n
for large n, we will apply the Limit Comparison Test comparing with the p-series with p = 2. Now,
n2 +4n
3n4 +9 n4 + 4n3 1
L= lim 1 = lim =
n→∞ n→∞ 3n4 + 9 3
n2

∞
n2 + 4n
The p-series with p = 2 converges and L exists; therefore, the series also converges.
n=1
3n4 + 9

∞
55. n−0.8
n=1
solution

∞ ∞
1
n−0.8 = 0.8
n=1 n=1
n
so that this is a divergent p-series.
108 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)


∞
56. (0.8)−n n−0.8
n=1
solution
∞ ∞ ∞
1.25n
(0.8)−n n−0.8 = (0.8−1 )n n−0.8 =
n=1 n=1 n=1
n0.8
1.25n
With an = n0.8
we have
   n 0.8
 an+1  = 1.25
n+1
n0.8
 an  (n + 1)0.8 · 1.25n = 1.25 n + 1
so that
 
an+1 
ρ = lim   = 1.25 > 1
n→∞ an 

Thus the original series diverges, by the Ratio Test.


∞
57. 4−2n+1
n=1
solution Observe

∞ 
∞ ∞  n
−2n+1 −2 n 1
4 = 4 · (4 ) = 4
n=1 n=1 n=1
16

is a geometric series with r = 1

16 ; therefore, this series converges.
∞
(−1)n−1
58. √
n=1
n
solution This is an alternating series with an = √1n . Because an forms a decreasing sequence which con-
∞
(−1)n−1
verges to zero, the series √ converges by the Alternating Series Test.
n=1
n
∞
1
59. sin 2
n=1
n
solution Here, we will apply the Limit Comparison Test, comparing with the p-series with p = 2. Now,
sin n12 sin u
L = lim 1
= lim =1
n→∞ u→0 u
n2

∞
1
where u = 1
n2
. The p-series with p = 2 converges and L exists; therefore, the series sin also converges.
n=1
n2
∞
1
60. (−1)n cos
n=1
n
solution Because
1
lim cos = cos 0 = 1  0
n→∞ n

∞
1
the general term in the series (−1)n cos does not tend toward zero; therefore, the series diverges by the
n=1
n
Divergence Test.
∞
(−2)n
61. √
n=1
n
solution Because
2n 2x 2 x ln 2 √
lim √ = lim √ = lim 1
= lim 2 x+1 x ln 2 = ∞  0
n→∞ n x→∞ x x→∞ 2 √ x x→∞

∞
(−2)n
the general term in the series √ does not tend toward zero; therefore, the series diverges by the
n=1
n
Divergence Test.
 S E C T I O N 10.5 The Ratio and Root Tests and Strategies for Choosing Tests (LT SECTION 11.5) 109

∞ 
 n n
62.
n=1
n + 12
solution Because the general term has the form of a function of n raised to the nth power, we might be
tempted to use the Root Test; however, the Root Test is inconclusive for this series. Instead, note
 −n ⎡ n/12 ⎤−12
12 ⎢⎢ ⎥⎥⎥
= lim ⎢⎢⎢⎣ 1 +
12 ⎥⎥⎦ = e−12  0
lim an = lim 1 +
n→∞ n→∞ n n→∞ n

Therefore, the series diverges by the Divergence Test.

Further Insights and Challenges


∞
√n
63. Proof of the Root Test Let an be a positive series, and assume that L = lim an exists.
n→∞
n=0
(a) Show that the series converges if L < 1. Hint: Choose R with L < R < 1 and show that an ≤ Rn for n

sufficiently large. Then compare with the geometric series Rn .
(b) Show that the series diverges if L > 1.
√
solution Suppose lim n an = L exists.
n→∞
1−L
(a) If L < 1, let  = . By the definition of a limit, there is a positive integer N such that
2
√
− ≤ n an − L ≤ 
for n ≥ N. From this, we conclude that
√n
0≤ an ≤ L + 
for n ≥ N. Now, let R = L + . Then
1−L L+1 1+1
R= L+ = < =1
2 2 2
and
√n
0≤ an ≤ R or 0 ≤ an ≤ Rn

∞ 
∞
for n ≥ N. Because 0 ≤ R < 1, the series Rn is a convergent geometric series, so the series an converges
n=N n=N

∞
by the Direct Comparison Test. Therefore, the series an also converges.
n=0
L−1
(b) If L > 1, let  = . By the definition of a limit, there is a positive integer N such that
2
√
− ≤ n an − L ≤ 
for n ≥ N. From this, we conclude that
√n
L− ≤ an
for n ≥ N. Now, let R = L − . Then
L−1 L+1 1+1
R= L− = > =1
2 2 2
and
√n
R≤ an or Rn ≤ an

∞ 
∞
for n ≥ N. Because R > 1, the series Rn is a divergent geometric series, so the series an diverges by
n=N n=N

∞
the Direct Comparison Test. Therefore, the series an also diverges.
n=0
110 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

64. Show that the Ratio Test does not apply, but verify convergence using the Direct Comparison Test for
the series
1 1 1 1 1
+ 2 + 3 + 4 + 5 + ···
2 3 2 3 2
solution The general term of the series is:
⎧1
⎪
⎪
⎪
⎪
⎪
⎨ 2n
n odd
an = ⎪
⎪
⎪
⎪
⎪ 1
⎩ n even
3n
First use the Ratio Test. If n is even,
1  n
an+1 2n+1 3n 1 3
= 1
= n+1 = ·
an 3n
2 2 2

whereas, if n is odd,
1  n
an+1 n+1 2n 1 2
= 3 1 = n+1 = ·
an 2n
3 3 3
 n  n
1 2 1 3 an+1
Since lim · = 0 and lim · = ∞, the sequence does not converge, and the Ratio Test is
n→∞ 3 3 n→∞ 2 2 an
inconclusive.
1
However, we have 0 ≤ an ≤ n for all n, so the series converges by comparison with the convergent
2
geometric series
∞
1
2n
n=1


∞
cn n!
65. Let , where c is a constant.
n=1
nn
(a) Prove that the series converges absolutely if |c| < e and diverges if |c| > e.
en n! √
(b) It is known that lim n+1/2 = 2π. Verify this numerically.
n→∞ n
(c) Use the Limit Comparison Test to prove that the series diverges for c = e.
solution
cn n!
(a) With an = nn ,
     −n
 an+1  = |c| (n + 1)! · n = |c| n n = |c| 1 + 1
n+1 n
 an  (n + 1)n+1 |c|n n! n+1 n
and
 
an+1 
ρ = lim   = |c|e−1
n→∞ an 


∞
cn n!
Thus, by the Ratio Test, the series converges when |c|e−1 < 1, or when |c| < e. The series diverges
n=1
nn
when |c| > e.
en n!
√
(b) The table below lists the value of nn+1/2
for several increasing values of n. Since 2π = 2.506628275, the
numerical evidence verifies that
en n! √
lim n+1/2
= 2π
n→∞ n

n 100 1000 10000 100000

e n!
n

nn+1/2
2.508717995 2.506837169 2.506649163 2.506630363
 S E C T I O N 10.6 Power Series (LT SECTION 11.6) 111


∞
en n!
(c) With c = e, the series becomes . Using the result from part (b),
n=1
nn

en n! √
nn en n!
L = lim √ = lim n+1/2 = 2π
n→∞ n n→∞ n
∞
√ ∞
en n!
Because the series n diverges by the Divergence Test and L > 0, we conclude that diverges by
n=1 n=1
nn
the Limit Comparison Test.

10.6 Power Series (LT Section 11.6)

Preliminary 
Questions
1. Suppose that an xn converges for x = 5. Must it also converge for x = 4? What about x = −3?

solution The power series an xn is centered at x = 0. Because the series converges for x = 5, the radius
of convergence must be at least 5 and the series converges absolutely at least for the interval |x| < 5. Both
x = 4 and x = −3 are inside this interval, so the series converges for x = 4 and for x = −3.

2. Suppose that an (x − 6)n converges for x = 10. At which of the points (a)–(d) must it also converge?

(a) x = 8 (b) x = 11 (c) x = 3 (d) x = 0

solution The given power series is centered at x = 6. Because the series converges for x = 10, the radius
of convergence must be at least |10 − 6| = 4 and the series converges absolutely at least for the interval
|x − 6| < 4, or 2 < x < 10.
(a) x = 8 is inside the interval 2 < x < 10, so the series converges for x = 8.
(b) x = 12 is not inside the interval 2 < x < 10, so the series may or may not converge for x = 12.
(c) x = 2 is an endpoint of the interval 2 < x < 10, so the series may or may not converge for x = 2.
(d) x = 0 is not inside the interval 2 < x < 10, so the series may or may not converge for x = 0.
3. What is the radius of convergence of F(3x) if F(x) is a power series with radius of convergence R = 12?
solution If the power series F(x) has radius of convergence R = 12, then the power series F(3x) has radius
of convergence R = 12
3 = 4.
∞
4. The power series F(x) = nxn has radius of convergence R = 1. What is the power series expansion of
n=1
F  (x) and what is its radius of convergence?
solution We obtain the power series expansion for F  (x) by differentiating the power series expansion for
F(x) term-by-term. Thus,

∞
F  (x) = n2 xn−1
n=1

The radius of convergence for this series is R = 1, the same as the radius of convergence for the series
expansion for F(x).

Exercises
∞
xn
1. Use the Ratio Test to determine the radius of convergence R of . Does it converge at the endpoints
n=0
2n
x = ±R?
xn
solution With an = 2n ,
   
 an+1  = |x| |x| an+1  |x|
n+1
2n
 an  · n = and ρ = lim  =
2n+1 |x| 2 n→∞ an  2
|x| |x|
By the Ratio Test, the series converges when ρ = 2 < 1, or |x| < 2, and diverges when ρ = 2 > 1, or |x| > 2.
The radius of convergence is therefore R = 2.
112 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)


∞
xn
2. Use the Ratio Test to show that √ n has radius of convergence R = 2. Then determine whether it
n=1
n2
converges at the endpoints R = ±2.
n
solution With an = √x n ,
n2
  √ n  
 an+1  = |x|n+1
·
n2
=
|x|
·
n
ρ = lim 
an+1  |x|
 = ·1=
|x|
 an  √ and
n + 12n+1 |x|n 2 n+1 n→∞ an 2 2
|x| |x|
By the Ratio Test, the series converges when ρ = 2 < 1, or |x| < 2, and diverges when ρ = 2 > 1, or |x| > 2.
The radius of convergence is therefore R = 2.
For the endpoint x = 2, the series becomes
∞
2n ∞
1
√ n = √
n=1
n2 n=1
n

which is a divergent p-series. For the endpoint x = −2, the series becomes
∞
(−2)n  (−1)n
∞
√ n = √
n=1
n2 n=1
n

This alternating series converges by the Alternating Series Test, but its associated positive series is a divergent
p-series. Thus, the series for x = −2 is conditionally convergent.
3. Show that the power series (a)–(c) have the same radius of convergence. Then show that (a) diverges at
both endpoints, (b) converges at one endpoint but diverges at the other, and (c) converges at both endpoints.
∞
xn ∞
xn ∞
xn
(a) (b) (c)
n=1
3n n=1
n3n n=1
n2 3n

solution
xn
(a) With an = 3n ,
   n+1 n   
 an+1   x 3  1 |x|
= lim  n+1 · n  = lim   = |x| lim =
x
ρ = lim  
n→∞ an  n→∞  3 x  n→∞ 3 n→∞ 3 3
Thus ρ < 1 when |x| < 3, so the radius of convergence is R = 3. For the endpoint x = 3, the series becomes

∞
3n 
∞
= 1
n=1
3n n=1

which diverges by the Divergence Test. For the endpoint x = −3, the series becomes

∞
(−3)n 
∞
= (−1)n
n=1
3n n=1

which also diverges by the Divergence Test.

n
(b) With an = n3x n ,
     
 an+1   xn+1 n3n   nx  n |x|
ρ = lim   = lim  · n  = lim   = |x| lim =
n→∞ an  n→∞  (n + 1)3n+1 x  n→∞ 3(n + 1)  n→∞ 3n + 3 3
Thus ρ < 1 when |x| < 3, so the radius of convergence is R = 3. For the endpoint x = 3, the series becomes
∞
3n ∞
1
n
=
n=1
n3 n=1
n

which is the divergent harmonic series. For the endpoint x = −3, the series becomes

∞
(−3)n 
∞
(−1)n
=
n=1
n3n n=1
n

which converges by the Alternating Series Test.

 S E C T I O N 10.6 Power Series (LT SECTION 11.6) 113

xn
(c) With an = n2 3n
,
     
 an+1   xn+1 n2 3n   n2 x  n2 |x|
ρ = lim   = lim  ·  = lim   = |x| lim =
n→∞ an n→∞  (n + 1)2 3n+1 x n  n→∞  3(n + 1) 2  n→∞ 3n + 6n + 3
2 3

Thus ρ < 1 when |x| < 3, so the radius of convergence is R = 3. For the endpoint x = 3, the series becomes
∞
3n ∞
1
2 n
=
n=1
n3 n=1
n2

which is a convergent p-series. For the endpoint x = −3, the series becomes

∞
(−3)n 
∞
(−1)n
=
n=1
n2 3n n=1
n2

which converges by the Alternating Series Test.

4. Repeat Exercise 3 for the following series:

∞
(x − 5)n ∞
(x − 5)n 
∞
(x − 5)n
(a) (b) (c)
n=1
9n n=1
n9n n=1
n2 9n

solution
(x−5)n
(a) With an = 9n ,
       
 an+1   (x − 5)n+1 9n   x − 5  =  x − 5 
ρ = lim   = lim  ·  = lim
n→∞ an n→∞  9n+1 (x − 5)n  n→∞  9   9 

Then ρ < 1 when |x − 5| < 9, or −4 < x < 14. So the radius of convergence is R = 9 centered at x = 5. For
the endpoint x = 14, the series becomes

∞
(14 − 5)n 
∞
= 1
n=1
9n n=1

which diverges by the Divergence Test. For the endpoint x = −4, the series becomes

∞
(−4 − 5)n 
∞
= (−1)n
n=1
9n n=1

which also diverges by the Divergence Test.

n
(b) With an = (x−5)
n9n ,
     
 an+1   (x − 5)n+1 n9n   n(x − 5)  = |x − 5| lim n = |x − 5|
ρ = lim   = lim  ·  = lim
n→∞ an n→∞  (n + 1)9n+1 (x − 5)n  n→∞  9(n + 1)  n→∞ 9n + 9 9

Then ρ < 1 when |x − 5| < 9, or −4 < x < 14. So the radius of convergence is R = 9 centered at x = 5. For
the endpoint x = 14, the series becomes

∞
(14 − 5)n 
∞
1
=
n=1
n9n n=1
n

which is the divergent harmonic series. For the endpoint x = −4, the series becomes

∞
(−4 − 5)n 
∞
(−1)n
=
n=1
n9n n=1
n

which converges by the Alternating Series Test.

n
(c) With an = (x−5)
n2 9n
,
     2 
 an+1   (x − 5)n+1 n2 9n   n (x − 5)  n2 |x − 5|
ρ = lim   = lim  ·  = lim   = |x − 5| lim 2 =
n→∞ an n→∞  (n + 1) 9
2 n+1 (x − 5)  n→∞  9(n + 1) 
n 2 n→∞ 9n + 18n + 9 9
114 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

Then ρ < 1 when |x − 5| < 9, or −4 < x < 14. So the radius of convergence is R = 9 centered at x = 5. For
the endpoint x = 14, the series becomes

∞
(14 − 5)n ∞
1
=
n=1
n2 9n n=1
n2

which is a convergent p-series. For the endpoint x = −4, the series becomes

∞
(−4 − 5)n 
∞
(−1)n
=
n=1
n2 9n n=1
n2

which converges by the Alternating Series Test.


∞
5. Show that nn xn diverges for all x  0.
solution Withn=0 an = nn xn ,
     n
 an+1   (n + 1)n+1 xn+1  (n + 1)n+1 1
ρ = lim   = lim   = |x| n→∞
lim = |x| lim 1 + (n + 1) = ∞ if x  0
n→∞ an  n→∞  nn xn nn n→∞ n
Then ρ < 1 only when x = 0, so the radius of convergence is R = 0. In other words, the power series
converges only for x = 0.

∞
6. For which values of x does n!xn converge?
n=0

solution With an = n!xn , and assuming x  0,

   
 an+1   (n + 1)!xn+1 
ρ = lim   = lim  lim |(n + 1)x| = ∞ if x  0
 = n→∞
n→∞ an  n→∞  n!xn
ρ < 1 only if x = 0, so that the radius of convergence is R = 0. In other words, the power series converges
only for x = 0.
 ∞
x2n √
7. Use the Ratio Test to show that n
has radius of convergence R = 3.
n=0
3
x2n
solution With an = n ,
3
   2(n+1)   2   2 
 an+1   x 3n  x  x 
ρ = lim   = lim  n+1 · 2n  = lim   =  
n→∞ an n→∞  3 x  n→∞  3   3 
√ √
Then ρ < 1 when |x2 | < 3, or |x| < 3, so the radius of convergence is R = 3.
 x3n+1
∞
8. Show that has radius of convergence R = 4.
n=0
64n
x3n+1
solution With an = ,
64n
   3(n+1)+1   3   3 
 an+1   x 64n  x  x 
ρ = lim   = lim  · 3n+1  = lim   =  
n→∞ an n→∞  64 n+1 x  n→∞  64   64 

Then ρ < 1 when |x|3 < 64 or |x| < 4, so the radius of convergence is R = 4.
In Exercises 9–34, find the interval of convergence.

∞
9. nxn
n=0

solution With an = nxn ,

   
an+1   (n + 1)xn+1  n+1
ρ = lim   = lim   = |x| n→∞
lim = |x|
n→∞ an n→∞  nxn n
Then ρ < 1 when |x| < 1, so that R = 1 and the series converges absolutely on the interval −1 < x < 1.

∞
For the endpoint x = 1, the series becomes n, which diverges by the Divergence Test. For the endpoint
n=0
 S E C T I O N 10.6 Power Series (LT SECTION 11.6) 115


∞ 
∞
x = −1, the series becomes (−1)n n, which also diverges by the Divergence Test. Thus, the series nxn
n=1 n=0
converges for −1 < x < 1 and diverges elsewhere.
∞
2n n
10. x
n=1
n
2n xn
solution With an = n ,
   n+1 n+1 
 an+1   2 x n  2n
ρ = lim   = lim  · n n  = |x| lim = 2|x|
n→∞ an n→∞  n+1 2 x  n→∞ n + 1

Then ρ < 1 when |x| < 1

so the radius of convergence is 12 , and the series converges absolutely on the
2,
 ∞
1
interval − 12 < x < 12 . For the endpoint x = 12 , the series becomes , which is the divergent harmonic
n=1
n
∞
(−1)n
series. For the endpoint x = − 2 , the series becomes
1
, which converges by the Alternating Series
n=1
n
 ∞
2n n
Test. Thus, the series x converges for − 12 ≤ x < 12 and diverges elsewhere.
n=1
n

∞
x2n+1
11. (−1)n
n=1
2n n

x2n+1
solution With an = (−1)n ,
2n n
 2(n+1)+1   2   
 x 2n n   x n   x2 
ρ = lim  n+1 ·  = lim  · = 
n→∞  2 (n + 1) x2n+1  n→∞  2 n + 1   2 
√ √
Then ρ < 1 when |x| < 2, so the radius of convergence is R = 2, and the series converges √ absolutely√on
√ √ √ ∞
− 2 ∞
2
the interval − 2 < x < 2. For the endpoint x = − 2, the series becomes (−1)n = (−1)n+1 ,
n n
n=1
∞ √
n=1
√ 2
which converges by the Leibniz test. For the endpoint x = 2, the series becomes (−1)n which also
n=1
n
∞
x2n+1 √ √
converges by the Leibniz test. Thus the series (−1)n n converges for − 2 ≤ x ≤ 2 and diverges
n=1
2 n
elsewhere.
∞
n
12. (−1)n n x2n
n=0
4
n 2n
solution With an = (−1)n x ,
4n
     2   
 an+1   (n + 1)x2(n+1) 4n   x n + 1   x2 
ρ = lim   = lim  · 2n  = lim  · = 
n→∞ an  n→∞  4n+1 nx  n→∞  4 n  4

Then ρ < 1 when |x2 | < 4, or |x| < 2, so the radius of convergence is R = 2, and the series converges

∞
absolutely for −2 < x < 2. At both endpoints x = ±2, the series becomes (−1)n n, which diverges by the
n=0
∞
n
Divergence Test. Thus, the series (−1)n n x2n converges for −2 < x < 2 and diverges elsewhere.
n=0
4
∞
xn
13.
n=4
n5
xn
solution With an = n5
,
   n+1 
 an+1   x n5  (n + 1)5
ρ = lim   = lim  · n  = |x| lim = |x|
n→∞ an n→∞  (n + 1) 5 x  n→∞ n5
116 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

Then ρ < 1 when |x| < 1, so the radius of convergence is R = 1, and the series converges absolutely on
∞
1
−1 < x < 1. For the endpoint x = 1, the series becomes 5
, which is a convergent p-series. For the
n=4
n
 ∞
(−1)n
endpoint x = −1, the series becomes , which converges by the Alternating Series Test. Thus, the
n=4
n5
∞
xn
series converges for −1 ≤ x ≤ 1 and diverges elsewhere.
n=4
n5


∞
14. n7 xn
n=8

solution With an = n7 xn ,

       
 an+1   (n + 1)7 xn+1   n + 1 7 
ρ = lim   = lim  lim  x
 = n→∞  = |x|
n→∞ an  n→∞  n7 xn  n 

Then ρ < 1 when |x| < 1, so that the radius of convergence is R = 1, and the series converges absolutely on

∞
the intervale −1 < x < 1. For the endpoint x = 1, the series becomes n7 , which diverges by the Divergence
n=8

∞
test; for the endpoints x = −1, the series becomes (−1)n n7 , which also diverges by the Divergence test.
n=8

∞
Thus the series n7 xn converges for −1 < x < 1 and diverges elsewhere.
n=8

∞
xn
15.
n=0
(n!)2

xn
solution With an = (n!)2
,
   n+1 
 an+1   x (n!)2  (n!)2 1
ρ = lim   = lim   = |x| lim = |x| lim =0
n→∞ an n→∞  ((n + 1)!) x 
2 n n→∞ ((n + 1)!)2 n→∞ (n + 1)2

Then ρ < 1 for any x, so the radius of convergence is R = ∞ and the series converges absolutely for all x.

∞
8n
16. xn
n=0
n!

8n xn
solution With an = n! ,

   n+1 n+1   
 an+1   8 x n!  1 
ρ = lim   = lim  · n n  = lim 8x · =0
n→∞ an n→∞  (n + 1)! 8 x  n→∞ n + 1

ρ < 1 for all x, so the radius of convergence is R = ∞, and the series converges absolutely for all x.
∞
(2n)! n
17. x
n=0
(n!)3

(2n)!xn
solution With an = (n!)3
, and assuming x  0,
     
an+1   (2(n + 1))!xn+1 (n!)3   x (2n + 2)(2n + 1) 
ρ = lim   = lim  ·  = lim
n→∞ an n→∞  ((n + 1)!)3 (2n)!xn  n→∞  (n + 1)3 
   
 4n2 + 6n + 2   4n−1 + 6n−1 + 2n−3 
= lim  x 3 =
 n→∞ 
lim x =0
n→∞  n + 3n2 + 3n + 1  1 + 3n−1 + 3n−2 + n−3 

Then ρ < 1 for all x, so the radius of convergence is R = ∞, and the series converges absolutely for all x.
 S E C T I O N 10.6 Power Series (LT SECTION 11.6) 117


∞
4n
18. x2n−1
n=0
(2n + 1)!

4n x2n−1
solution With an = (2n+1)! ,
   n+1 2n+1   
 an+1   4 x (2n + 1)!   4x2 
ρ = lim   = lim  · n 2n−1  = lim  =0
n→∞ an  n→∞  (2n + 3)! 4 x  n→∞  (2n + 3)(2n + 2) 
Then ρ is always less than 1, and the series converges absolutely for all x.

∞
(−1)n xn
19. √
n=0 n2 + 1

(−1)n xn
solution With an = √ ,
n2 +1
   √  √
 an+1   (−1)n+1 xn+1 n2 + 1  n2 + 1 n2 + 1
ρ = lim   = lim   ·  = |x| lim  = |x| lim = |x|
n→∞ an n→∞  (n + 1)2 + 1 (−1) x 
n n n→∞ (n + 1)2 + 1 n→∞ n2 + 2n + 2
Then ρ < 1 when |x| < 1, so the radius of convergence is R = 1, and the series converges absolutely
∞
(−1)n
on the interval −1 < x < 1. For the endpoint x = 1, the series becomes √ , which converges
n=0 n2 + 1
∞
1
by the Alternating Series Test. For the endpoint x = −1, the series becomes √ , which diverges
n=0 n2 + 1
∞
(−1)n xn
by the Limit Comparison Test comparing with the divergent harmonic series. Thus, the series √
n=0 n2 + 1
converges for −1 < x ≤ 1 and diverges elsewhere.


∞
xn
20.
n=0
n4 +2

n
solution With an = n4x+2 ,
   
an+1   xn+1 n4 + 2  (n + 1)4 + 2 n4 + 4n3 + 6n2 + 4n + 3
ρ = lim   = lim  ·  = |x| lim = |x| lim = |x|
n→∞ an n→∞  (n + 1) + 2
4 x 
n n→∞ n +2
4 n→∞ n4 + 2
Then ρ < 1 when |x| < 1, so the radius of convergence is R = 1, and the series converges absolutely on
∞
1
the interval −1 < x < 1. For the endpoint x = 1, the series becomes 4+2
. Because n41+2 < n14 and
n=0
n
∞
1
the series is a convergent p-series, the endpoint series converges by the Comparison Test. For the
n=1
n4
 ∞
(−1)n
endpoint x = −1, the series becomes , which converges by the Alternating Series Test. Thus, the
n=0
n4 + 2
∞
xn
series converges for −1 ≤ x ≤ 1 and diverges elsewhere.
n=0
n4 + 2

∞
x2n+1
21.
n=15
3n + 1

2n+1
solution With an = 3n+1x
,
   2(n+1)+1   2n+3 
 an+1   x 3n + 1   x (3n + 1)  3n + 1
ρ = lim   = lim  · 2n+1  = lim  = |x|2 lim = |x|2
n→∞ an  n→∞  3(n + 1) + 1 x  n→∞  (3n + 4)x2n+1  n→∞ 3n + 4

Then ρ < 1 when |x| < 1, so the radius of convergence is R = 1, and the series converges absolutely on the
∞
1
interval −1 < x < 1. For the endpoint x = 1, the series becomes , which diverges by the Limit
n=15
3n +1
118 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

Comparison Test comparing with the divergent harmonic series. For the endpoint x = −1, the series becomes
∞
−1
, which also diverges by the Limit Comparison Test comparing with the divergent harmonic series.
n=15
3n +1
∞
x2n+1
Thus, the series converges for −1 < x < 1 and diverges elsewhere.
n=15
3n + 1


∞
xn
22.
n=9
n − 4 ln n

xn
solution With an = n−4 ln n ,

   
 an+1   xn+1 n − 4 ln n  n − 4 ln n 1 − 4 lnnn
ρ = lim  
 = lim  ·  = |x| lim = |x| lim
n→∞ an n→∞  (n + 1) − 4 ln(n + 1) xn  n→∞ n + 1 − 4 ln(n + 1) n→∞ 1 + 1 − 4 ln(n+1)
n n

Now, using L’Hôpital’s rule, we have

ln n 1/n 1
lim = lim = lim = 0
n→∞ n n→∞ 1 n→∞ n

ln(n + 1) 1/(n + 1) 1
lim = lim = lim =0
n→∞ n n→∞ 1 n→∞ n + 1

so that
1 − 4 lnnn
|x| lim = |x|
n→∞ 1+ 1
n − 4 ln(n+1)
n

Then ρ < 1 when |x| < 1, so the radius of convergence is R = 1, and the series converges absolutely on the
∞
1 1 1
interval −1 < x < 1. For the endpoint x = 1, the series becomes . Because > and
n=9
n − 4 ln n n − 4 ln n n
∞
1
is the divergent harmonic series, the endpoint series diverges by the Comparison Test. For the endpoint
n=9
n
∞
(−1)n
x = −1, the series becomes , which converges by the Alternating Series Test. Thus, the series
n=9
n − 4 ln n
∞
xn
converges for −1 ≤ x < 1 and diverges elsewhere.
n=9
n − 4 ln n

∞
xn
23.
n=2
ln n

xn
solution With an = ln n ,
   
 an+1   xn+1 ln n  ln n
ρ = lim   = lim  · n  = |x| lim
n→∞ an n→∞  ln(n + 1) x  n→∞ ln(n + 1)

Using L’Hôpital’s Rule, we have

ln n 1/n n+1
ρ = |x| lim = |x| lim = |x| lim = |x|
n→∞ ln(n + 1) n→∞ 1/(n + 1) n→∞ n
Then ρ < 1 when |x| < 1, so the radius of convergence is R = 1, and the series converges absolutely on the
∞
1  ∞
1
interval −1 < x < 1. For the endpoint x = 1, the series becomes . Because ln1n > 1n and is the
n=2
ln n n=2
n
divergent harmonic series, the endpoint series diverges by the Comparison Test. For the endpoint x = −1, the
∞
(−1)n ∞
xn
series becomes , which converges by the Alternating Series Test. Thus, the series converges
n=2
ln n n=2
ln n
for −1 ≤ x < 1 and diverges elsewhere.
 S E C T I O N 10.6 Power Series (LT SECTION 11.6) 119

∞
x3n+2
24.
n=2
ln n

x3n+2
solution With an = ln n ,

   3(n+1)+2   3n+5 
an+1  x ln n   x ln n 
ρ = lim 
ln n
 = lim  · 3n+2  = lim  · 3n+2  = |x3 | lim
n→∞ an n→∞  ln(n + 1) x  n→∞  ln(n + 1) x  n→∞ ln(n + 1)

Using L’Hôpital’s Rule, we have

ln n 1/n n+1
ρ = |x3 | lim = |x3 | lim = |x3 | lim = |x3 |
n→∞ ln(n + 1) n→∞ 1/(n + 1) n→∞ n

Then ρ < 1 when |x3 | = |x|3 < 1, so the radius of convergence is R = 1, and the series converges absolutely
∞
1
on the interval −1 < x < 1. For the endpoint x = 1, the series becomes . Because ln1n > 1n and
n=2
ln n
∞
1
is the divergent harmonic series, the endpoint series diverges by the Comparison Test. For the endpoint
n=2
n
∞
(−1)n
x = −1, the series becomes , which converges by the Alternating Series Test. Thus, the series
n=2
ln n
∞
xn
converges for −1 ≤ x < 1 and diverges elsewhere.
n=2
ln n


∞
25. n(x − 3)n
n=1

solution With an = n(x − 3)n ,

     
 an+1   (n + 1)(x − 3)n+1   n + 1 
ρ = lim   = lim  lim 
 = |x − 3| n→∞  = |x − 3|
n→∞ an  n→∞  n(x − 3)n n 

Then ρ < 1 when |x − 3| < 1, so the radius of convergence is R = 1 centered at 3, and the series converges
∞
absolutely on the interval 2 < x < 4. For the endpoint x = 4, the series becomes n, which diverges
n=1

∞
by the Divergence Test. For the endpoint x = 2, the series becomes (−1)n n, which also diverges by the
n=1

∞
Divergence Test. Thus, the series n(x − 3) converges for 2 < x < 4 and diverges elsewhere.
n

n=1


∞
(−5)n (x − 3)n
26.
n=1
n2

(−5)n (x−3)n
solution With an = n2
,

   
 an+1   (−5)n+1 (x − 3)n+1 n2 
ρ = lim   = lim  · 
n→∞ an  n→∞  (n + 1)2 (−5)n (x − 3)n 
 
 −5n2  5n2
= |x − 3| lim   = |x − 3| lim = 5|x − 3|
n→∞  (n + 1)2  n→∞ n2 + 2n + 1

Then ρ < 1 when 5|x − 3| < 1, so when |x − 3| < 15 . Thus the radius of convergence is 15 centered at x = 3,
5 < x < 5 . For the endpoint x = 5 , the series becomes
so the series converges absolutely on the interval 14 16 16
120 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)


∞
(−1)n
, which converges by the Alternating Series Test. For the endpoint x = 14
5 , the series becomes
n=1
n2
∞
1 ∞
(−5)n (x − 3)n
, which is a convergent p-series. Thus, the series converges for 14
5 ≤ x ≤ 16
5 and
n=1
n2 n=1
n2
diverges elsewhere.


∞
27. (−1)n n5 (x − 7)n
n=1

solution With an = (−1)n n5 (x − 7)n ,

     
 an+1   (−1)n+1 (n + 1)5 (x − 7)n+1   −(n + 1)5  (n + 1)5
ρ = lim   = lim   = |x − 7| lim   = |x − 7| lim = |x − 7|
n→∞ an  n→∞  (−1)n n5 (x − 7)n  n→∞  n5  n→∞ n5
Then ρ < 1 when |x − 7| < 1, so the radius of convergence is R = 1 centered at 7, and the series converges
∞
absolutely on the interval 6 < x < 8. For the endpoint x = 6, the series becomes n5 , which diverges
n=1

∞
by the Divergence Test. For the endpoint x = 8, the series becomes n 5
(−1) n , which also diverges by the
n=1

∞
Divergence Test. Thus, the series (−1)n n5 (x − 7)n converges for 6 < x < 8 and diverges elsewhere.
n=1


∞
28. 27n (x − 1)3n+2
n=0

solution With an = 27n (x − 1)3n+2 ,

   n+1 
 an+1   27 (x − 1)3n+5   
ρ = lim   = lim   = lim 27(x − 1)3  = |27(x − 1)3 |
n→∞ an n→∞  27 (x − 1)
n 3n+2  n→∞

Then ρ < 1 when |27(x − 1)3 | < 1, or when |(x − 1)3 | < 27 1
, so when |x − 1| < 13 . Thus the radius of
convergence is 3 , and the series converges absolutely when 3 < x < 43 . For the endpoint x = 23 , the series
1 2

  3n+2
1
∞ ∞
n −1
becomes 27 = (−1)n which diverges by the Divergence Test. For the endpoint x = 43 ,
n=0
3 9 n=0
  3n+2
1
∞ ∞
1
the series becomes 27n = 1, which also diverges by the Divergence Test. Thus the series
n=0
3 9 n=0
∞
27n (x − 1)3n+2 converges for 23 < x < 43 and diverges elsewhere.
n=0

∞
2n
29. (x + 3)n
n=1
3n

n n
solution With an = 2 (x+3)
3n ,
   n+1   
an+1   2 (x + 3)n+1   6n  = 2|x + 3|
ρ = lim 
3n
 = lim  · n  = |x + 3| lim
n→∞ an n→∞  3(n + 1) 2 (x + 3)n  n→∞  3n + 3 

Then ρ < 1 when 2|x + 3| < 1, or when |x + 3| < 12 . Thus the radius of convergence is 12 centered at x = −3,
so the series converges absolutely on the interval − 27 < x < − 52 . For the endpoint x = − 52 , the series becomes
∞
1
, which diverges because it is a multiple of the divergent harmonic series. For the endpoint x = − 72 , the
n=1
3n
∞
(−1)n  ∞
2n
series becomes , which converges by the Alternating Series Test. Thus, the series (x + 3)n
n=1
3n n=1
3n
converges for − 72 ≤ x < − 52 and diverges elsewhere.
 S E C T I O N 10.6 Power Series (LT SECTION 11.6) 121


∞
(x − 4)n
30.
n=0
n!

(x−4)n
solution With an = n! ,

     
 an+1   (x − 4)n+1 n!  (x − 4) 1  = 0
ρ = lim   = lim  ·  = lim
n→∞ an n→∞  (n + 1)! (x − 4)n  n→∞  n


∞
(x − 4)n
Thus ρ < 1 for all x, so the radius of convergence is infinite, and converges for all x.
n=0
n!


∞
(−5)n
31. (x + 10)n
n=0
n!

(−5)n
solution With an = n! (x + 10)n ,
     
 an+1   (−5)n+1 (x + 10)n+1 n!   1 
ρ = lim   = lim  · 
 = lim 5(x + 10)  = 0
n→∞ an  n→∞  (n + 1)! (−5)n (x + 10)n  n→∞  n


∞
(−5)n
Thus ρ < 1 for all x, so the radius of convergence is infinite, and (x + 10)n converges for all x.
n=0
n!


∞
32. n! (x + 5)n
n=10

solution With an = n! (x + 5)n ,

   
 an+1   (n + 1)! (x + 5)n+1 
ρ = lim   = lim  lim |(n + 1)(x + 5)| = ∞ if x  −5
 = n→∞
n→∞ an  n→∞  n! (x + 5)n

Thus ρ < 1 only when x + 5 = 0, so the radius of convergence is zero, and the series converges absolutely
only for x = −5.


∞
33. en (x − 2)n
n=12

solution With an = en (x − 2)n ,

   n+1 
 an+1   e (x − 2)n+1 
ρ = lim   = lim  n lim |e(x − 2)| = |e(x − 2)|
 = n→∞
n→∞ an  n→∞  e (x − 2)n 

Thus ρ < 1 when |e(x − 2)| < 1, or when |x − 2| < e−1 . So the radius of convergence is e−1 centered at x = 2,
and the series converges absolutely on the interval 2 − e−1 < x < 2 + e−1 . For the endpoint x = 2 + e−1 ,
∞
the series becomes 1, which diverges by the Divergence Test. For the endpoint x = 2 − e−1 , the series
n=12

∞ 
∞
becomes n
(−1) , which also diverges by the Divergence Test. Thus, the series en (x − 2)n converges for
n=12 n=12
2 − e−1 < x < 2 + e−1 and diverges elsewhere.

∞
(x + 4)n
34.
n=2
(n ln n)2

(x+4)n
solution With an = (n ln n)2
,
122 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

   
 an+1   (x + 4)n+1 (n ln n)2 
ρ = lim   = lim  · 
n→∞ an  n→∞  ((n + 1) ln(n + 1))2 (x + 4)n 
 2
(n ln n)2 n ln n
= |x + 4| lim = |x + 4| lim ·
n→∞ ((n + 1) ln(n + 1))2 n→∞ n + 1 ln(n + 1)

By L’Hôpital’s rule,
ln n 1/n n+1
lim = lim = lim =1
n→∞ ln(n + 1) n→∞ 1/(n + 1) n→∞ n
so that both factors in the limit above approach 1 as n → ∞. It follows that
 2
n ln n
ρ = |x + 4| lim · = |x + 4|
n→∞ n + 1 ln(n + 1)

Then ρ < 1 when |x + 4| < 1, so the radius of convergence is 1 centered at −4, and the series converges
∞
1
absolutely on the interval −5 < x < −3. For the endpoint x = −3, the series becomes , which
n=2
(n ln n)2
 ∞
1
converges by the Limit Comparison Test comparing with the convergent p-series 2
. For the endpoint
n=2
n
∞
(−1)n
x = −5, the series becomes , which converges by the Alternating Series Test. Thus, the series
n=2
(n ln n)2
∞
(x + 4)n
converges for −5 ≤ x ≤ −3 and diverges elsewhere.
n=2
(n ln n)2

In Exercises 35–40, use Eq. (2) to expand the function in a power series with center c = 0 and determine the
interval of convergence.
1
35. f (x) =
1 − 3x

solution Substituting 3x for x in Eq. (2), we obtain

1  ∞  ∞
= (3x)n = 3n xn
1 − 3x n=0 n=0

This series is valid for |3x| < 1, or |x| < 13 .

1
36. f (x) =
1 + 3x

solution Substituting −3x for x in Eq. (2), we obtain

1  ∞  ∞
= (−3x)n = (−3)n xn
1 + 3x n=0 n=0

This series is valid for | − 3x| < 1, or |x| < 13 .

1
37. f (x) =
3−x

solution First write

1 1 1
= ·
3−x 3 1− x
3
x
Substituting 3 for x in Eq. (2), we obtain
∞  n
 ∞
1 x xn
= =
1− x
3 n=0
3 n=0
3n
 S E C T I O N 10.6 Power Series (LT SECTION 11.6) 123

Thus,

1  xn  xn
∞ ∞
1
= =
3 − x 3 n=0 3n n=0 3n+1

This series is valid for |x/3| < 1, or |x| < 3.

1
38. f (x) =
4 + 3x

solution First write

1 1 1
= ·
4 + 3x 4 1 + 3x
4

Substituting − 3x
4 for x in Eq. (2), we obtain

∞  n  ∞ n n
1 3x n3 x
= − = (−1)
1 + 3x
4 n=0
4 n=0
4n

Thus,

1 3n xn 
∞ ∞
1 3n xn
= (−1)n n = (−1)n n+1
4 + 3x 4 n=0 4 n=0
4

This series is valid for | − 3x/4| < 1, or |x| < 43 .

1
39. f (x) =
1 − x3

solution Substituting x3 for x in Eq. (2), we obtain

1  ∞
n 
∞
= x3 = x3n
1 − x3 n=0 n=0

This series is valid for |x3 | < 1, or |x| < 1.

1
40. f (x) =
1 − x4

solution Substituting x4 for x in Eq. (2), we obtain

1  ∞
n 
∞
= x4 = x4n
1−x 4
n=0 n=0

This series is valid for |x4 | < 1, or |x| < 1.

3x2
41. Differentiate the power series in Exercise 39 to obtain a power series for g(x) = .
(1 − x3 )2
solution The general term in Exercise 39 is x3n , whose derivative is 3nx3n−1 . The term for n = 0 is a
1
constant, so it vanishes when we take the derivative. The derivative of is g(x), so the power series
1 − x3
about c = 0 for g(x) is


∞
3nx3n−1
n=1

This series is valid for |x| < 1.

124 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

4x3
42. Differentiate the power series in Exercise 40 to obtain a power series for g(x) = .
(1 − x4 )2
solution The general term in Exercise 40 is x4n , whose derivative is 4nx4n−1 . The term for n = 0 is a
1
constant, so it vanishes when we take the derivative. The derivative of is g(x), so the power series
1 − x4
about c = 0 for g(x) is

∞
4nx4n−1
n=1

This series is valid for |x| < 1.

1
43. (a) Divide the power series in Exercise 41 by 3x2 to obtain a power series for h(x) = and use
(1 − x3 )2
the Ratio Test to show that the radius of convergence is 1.
(b) Another way to obtain a power series for h(x) is to square the power series for f (x) in Exercise 39. By
multiplying term by term, determine the terms up to degree 9 in the resulting power series for ( f (x))2 and
show that they match the terms in the power series for h(x) found in part (a).
solution
1
(a) Dividing the series in Exercise 41 by 3x2 gives the power series for :
(1 − x3 )2

∞
3nx3n−1 
∞ 
∞
= nx3n−3 = (n + 1)x3n
n=1
3x2 n=1 n=0

By the Ratio Test,

 
an+1  (n + 2)x3(n+1) n+2
lim   = lim = x3 lim = x3
n→∞ an n→∞ (n + 1)x3n n→∞ n + 1

so the radius of convergence is 1.

(b) Multiplying the first few terms of the power series in Exercise 41 by themselves gives

(1 + x3 + x6 + x9 + · · · )(1 + x3 + x6 + x9 + · · · ) = 1 + 2x3 + 3x6 + 4x9 + · · ·

which matches terms of the form (n + 1)x3n for n = 0, 1, 2, 3.

1
44. (a) Divide the power series in Exercise 42 by 4x3 to obtain a power series for h(x) = and use
(1 − x4 )2
the Ratio Test to show that the radius of convergence is 1.
(b) Another way to obtain a power series for h(x) is to square the power series for f (x) in Exercise 40. By
multiplying term by term, determine the terms up to degree 12 in the resulting power series for ( f (x))2 and
show that they match the terms in the power series for h(x) found in part (a).
solution
1
(a) Dividing the series in Exercise 42 by 4x3 gives the power series for :
(1 − x4 )2

∞
4nx4n−1 
∞ 
∞
= nx4n−4 = (n + 1)x4n
n=1
4x3 n=1 n=0

By the Ratio Test,

 
an+1  (n + 2)x4(n+1) n+2
lim   = lim = x4 lim = x4
n→∞ an n→∞ (n + 1)x4n n→∞ n + 1

so the radius of convergence is 1.

(b) Multiplying the first few terms of the power series in Exercise 42 by themselves gives

(1 + x4 + x8 + x12 + · · · )(1 + x4 + x8 + x12 + · · · ) = 1 + 2x4 + 3x8 + 4x12 + · · ·

which matches terms of the form (n + 1)x4n for n = 0, 1, 2, 3.

 S E C T I O N 10.6 Power Series (LT SECTION 11.6) 125

45. Use the equalities

1 1 − 31
= =
1 − x −3 − (x − 4) 1 + x−4 3

to show that for |x − 4| < 3,

1  ∞
(x − 4)n
= (−1)n+1 n+1
1 − x n=0 3

solution Substituting − x−4

3 for x in Eq. (2), we obtain

∞  n  ∞
1 x−4 (x − 4)n
 = − = (−1)n
1+ x−4
n=0
3 n=0
3n
3

Thus,

1 (x − 4)n 
∞ ∞
1 (x − 4)n
=− (−1)n = (−1)n+1 n+1
1−x 3 n=0 3n
n=0
3

This series is valid for | − 3 |

x−4
< 1, or |x − 4| < 3.

46. Use the method of Exercise 45 to expand 1/(1 − x) in power series with centers c = 2 and c = −2.
Determine the interval of convergence.

solution For c = 2, write

1 1 1
= =−
1 − x −1 − (x − 2) 1 + (x − 2)

Substituting −(x − 2) for x in Eq. (2), we obtain

1  ∞  ∞
= (−(x − 2))n = (−1)n (x − 2)n
1 + (x − 2) n=0 n=0

Thus,

1 ∞ ∞
= − (−1)n (x − 2)n = (−1)n+1 (x − 2)n
1−x n=0 n=0

This series is valid for | − (x − 2)| < 1, or |x − 2| < 1.

For c = −2, write

1 1 1 1
= = ·
1 − x 3 − (x + 2) 3 1 − x+2
3

x+2
Substituting 3 for x in Eq. (2), we obtain

∞  n  ∞
1 x+2 (x + 2)n
= =
1 − x+2
3 n=0
3 n=0
3n

Thus,

1  (x + 2)n  (x + 2)n
∞ ∞
1
= =
1 − x 3 n=0 3n n=0
3n+1

This series is valid for | x+2

3 | < 1, or |x + 2| < 3.
126 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

47. Use the method of Exercise 45 to expand 1/(4 − x) in a power series with center c = 5. Determine the
interval of convergence.
solution First write

1 1 1
= =−
4 − x −1 − (x − 5) 1 + (x − 5)
Substituting −(x − 5) for x in Eq. (2), we obtain

1  ∞  ∞
= (−(x − 5))n = (−1)n (x − 5)n
1 + (x − 5) n=0 n=0

Thus,

1 ∞ ∞
= − (−1)n (x − 5)n = (−1)n+1 (x − 5)n
4−x n=0 n=0

This series is valid for | − (x − 5)| < 1, or |x − 5| < 1.

48. Find a power series that converges only for x in [2, 6).
6+2
solution The power series must be centered at c = = 4, with radius of convergence R = 2.
2

∞
(x − 4)n
n=1
n2n

(x−4)n
With an = n2n ,Consider the following series:
   
 an+1   (x − 4)n+1 n2n  n 1
ρ = lim   = lim  ·  = |x − 4| n→∞
lim = |x − 4|
n→∞ an  n→∞  (n + 1)2n+1 (x − 4)n  2(n + 1) 2

Thus ρ < 1 only when 12 |x − 4| < 1, or |x − 4| < 2, so the radius of convergence is 2 centered at 4, and
the series converges absolutely on the interval 2 < x < 6. At the endpoint x = 6, the series becomes
∞
(6 − 4)n  1
∞
= , which is the divergent harmonic series. For the endpoint x = 2, the series becomes
n=1
n·2n n=1
n
∞
(2 − 4)n  (−1)n
∞

n = , which converges by the Alternating Series Test. Therefore, the series converges for
n=1
n·2 n=1
n
2 ≤ x < 6, as desired.
49. Apply integration to the expansion

1  ∞
= (−1)n xn = 1 − x + x2 − x3 + · · ·
1 + x n=0

to prove that for −1 < x < 1,


∞
(−1)n−1 xn x2 x3 x4
ln(1 + x) = = x− + − + ···
n=1
n 2 3 4

solution To obtain the first expansion, substitute −x for x in Eq. (2):

1  ∞  ∞
= (−x)n = (−1)n xn
1 + x n=0 n=0

This expansion is valid for | − x| < 1, or −1 < x < 1.

Upon integrating both sides of the above equation, we find
⎛∞ ⎞
⎜⎜⎜ ⎟
ln(1 + x) =
dx
= ⎜⎜⎝ (−1)n xn ⎟⎟⎟⎟⎠ dx
1+x n=0
 S E C T I O N 10.6 Power Series (LT SECTION 11.6) 127

Integrating the series term-by-term then yields


∞
xn+1
ln(1 + x) = C + (−1)n
n=0
n+1

To determine the constant C, set x = 0. Then 0 = ln(1 + 0) = C. Finally,

∞
xn+1 
∞
xn
ln(1 + x) = (−1)n = (−1)n−1
n=0
n + 1 n=1 n

50. Use the result of Exercise 49 to prove that

3 1 1 1 1
ln = − + − + ···
2 2 2 · 22 3 · 23 4 · 24
Use your knowledge of alternating series to find an N such that the partial sum S N approximates ln 32 to
within an error of at most 10−3 . Confirm using a calculator to compute both S N and ln 32 .
solution In the previous exercise we found that
∞
xn+1
ln(1 + x) = (−1)n
n=0
n+1

Setting x = 1
2 yields:
 n
3  
∞ 1 ∞
n−1 2 (−1)n−1 1 1 1 1
ln = (−1) = = − + − + ···
2 n=1 n n=1
n2n 2 2 · 22 3 · 23 4 · 24

Note that the series for ln 32 is an alternating series with an = n21n . The error in approximating ln 32 by the
partial sum S N is therefore bounded by
 
ln 3 − S  < a 1
 2 N N+1 =
(N + 1)2N+1
To obtain an error of at most 10−3 , we must find an N such that
1
< 10−3 or (N + 1)2N+1 > 1000
(N + 1)2N+1
For N = 6, (N + 1)2N+1 = 7 · 27 = 896 < 1000, but for N = 7, (N + 1)2N+1 = 8 · 28 = 2048 > 1000; hence,
the smallest value for N is N = 7. The corresponding approximation is
1 1 1 1 1 1 1
S7 = − + − + − + = 0.405803571
2 2 · 22 3 · 23 4 · 24 5 · 25 6 · 26 7 · 27
Now, ln 32 = 0.405465108, so
 
ln 3 − S  = 3.385 × 10−4 < 10−3
 2 7

51. Let F(x) = (x + 1) ln(1 + x) − x.

(a) Apply integration to the result of Exercise 49 to prove that for −1 < x < 1,

∞
xn+1
F(x) = (−1)n+1
n=1
n(n + 1)

(b) Evaluate at x = 1
2 to prove
3 3 1 1 1 1 1
ln − = − + − + ···
2 2 2 1 · 2 · 22 2 · 3 · 23 3 · 4 · 24 4 · 5 · 25
(c) Use a calculator to verify that the partial sum S 4 approximates the left-hand side with an error no greater
than the term a5 of the series.
128 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

solution
(a) Note that

ln(x + 1) dx = (x + 1) ln(x + 1) − x + C

Then integrating both sides of the result of Exercise 49 gives


∞
(−1)n−1 xn
(x + 1) ln(x + 1) − x = ln(x + 1) dx = dx
n=1
n
For −1 < x < 1, which is the interval of convergence of the series in Exercise 49, therefore, we can integrate
term by term to get
∞
(−1)n−1 ∞
(−1)n−1 xn+1 ∞
xn+1
(x + 1) ln(x + 1) − x = xn dx = · +C = (−1)n+1 +C
n=1
n n=1
n n+1 n=1
n(n + 1)
(noting that (−1)n−1 = (−1)n+1 ). To determine C, evaluate both sides at x = 0 to get
0 = ln 1 − 0 = 0 + C
so that C = 0 and we get finally

∞
xn+1
(x + 1) ln(x + 1) − x = (−1)n+1
n=1
n(n + 1)

(b) Evaluating the result of part(a) at x = 1

2 gives

3 3 1 
∞
1
ln − = (−1)n+1
2 2 2 n=1 n(n + 1)2n+1
1 1 1 1
= − + − + ...
1 · 2 · 22 2 · 3 · 23 3 · 4 · 24 4 · 5 · 25
(c)
1 1 1 1
S4 = − + − = 0.1078125
1·2·22 2·3·2 3 3·4·2 4 4 · 5 · 25
1
a5 = ≈ 0.0005208
5 · 6 · 26
3 3 1
ln − ≈ 0.10819766
2 2 2
and
 
S − 3 ln 3 − 1  ≈ 0.0003852 < a
 4 2 2 2 5

52. Prove that for |x| < 1,

dx x5 x9
= A+x− + − ···
x4 +1 5 9
$ 1/2
Use the first two terms to approximate 0 dx/(x4 + 1) numerically. Use the fact that you have an alternating
series to show that the error in this approximation is at most 0.00022.
solution Substitute −x4 for x in Eq. (2) to get

1  ∞  ∞
= (−x4 )n = (−1)n x4n
1+x 4
n=0 n=0

This is valid for |x| < 1, so for x in that range we can integrate the right-hand side term by term to get

1  ∞ 
∞
x4n+1
dx = (−1)n x4n dx = (−1)n +C
1+x 4
n=0 n=0
4n + 1
5 9 13
x x x
= x− + − + ··· +C
5 9 13
 S E C T I O N 10.6 Power Series (LT SECTION 11.6) 129

Using the first two terms, we have

1/2
1 1 1 79
dx ≈ − 5 = = 0.49375
0 1+x 4 2 2 · 5 160
Since this is an alternating series, the error in the approximation is bounded by the first unused term, so by
1 1
= ≈ 0.000217 < 0.00022
29 ·9 4608
53. Use the result of Example 7 to show that
x2 x4 x6 x8
F(x) = − + − + ···
1·2 3·4 5·6 7·8
is an antiderivative of f (x) = tan−1 x satisfying F(0) = 0. What is the radius of convergence of this power
series?
solution For −1 < x < 1, which is the interval of convergence for the power series for arctangent, we can
integrate term-by-term, so integrate that power series to get

∞
(−1)n x2n+1  ∞
x2n+2
F(x) = tan−1 x dx = dx = (−1)n
n=0
2n + 1 n=0
(2n + 1)(2n + 2)
2 4
x x x6 x8
= − + − + ··· +C
1·2 3·4 5·6 7·8
If we assume F(0) = 0, then we have C = 0. The radius of convergence of this power series is the same as
that of the original power series, which is 1.
54. Verify that function F(x) = x tan−1 x − 12 ln(x2 + 1) is an antiderivative of f (x) = tan−1 x satisfying
F(0) = 0. Then use the result of Exercise 53 with x = √13 to show that

π 1 4 1 1 1 1
√ − ln = − + − + ···
6 3 2 3 1 · 2(3) 3 · 4(3 ) 5 · 6(3 ) 7 · 8(34 )
2 3

Use a calculator to compare the value of the left-hand side with the partial sum S 4 of the series on the right.
solution We have
x 1 1 x x
F  (x) = tan−1 x + − · 2 · 2x = tan−1 x + − = tan−1 x
1+x 2 2 x +1 1+x 2 1 + x2
so that F(x) is an antiderivative of tan−1 x, and clearly F(0) = 0. So applying Exercise 53 for this F, and
setting x = √13 , gives
 
1 1 1 1 π 1 4
√ tan−1 √ − ln + 1 = √ − ln
3 3 2 3 6 3 2 3
√ √ √ √
(1/ 3)2 (1/ 3)4 (1/ 3)6 (1/ 3)8
= − + − + ...
1·2 3·4 5·6 7·8
1 1 1 1
= − + − + ...
1 · 2(3) 3 · 4(32 ) 5 · 6(33 ) 7 · 8(34 )
Now, we have
1 1 1 1 3593
S4 = − + − = ≈ 0.1548215
1 · 2(3) 3 · 4(3 ) 5 · 6(3 ) 7 · 8(3 ) 22680
2 3 4

π 1 4
√ − ln ≈ 0.158459
6 3 2 3
so the two differ by less than 0.00004.
∞
n
55. Evaluate . Hint: Use differentiation to show that
n=1
2n

∞
(1 − x)−2 = nxn−1 (for |x| < 1)
n=1
130 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

solution Differentiate both sides of Eq. (2) to obtain

1  ∞
= nxn−1
(1 − x)2 n=1

Setting x = 1
2 then yields
∞
n 1
n−1
= 2
=4
n=1
2 1 − 12
Divide this equation by 2 to obtain
∞
n
n
=2
n=1
2

56. Use the power series for (1 + x2 )−1 and differentiation to prove that for |x| < 1,

2x  ∞
= (−1)n−1 (2n)x2n−1
(x2 + 1)2 n=1

solution From Exercise 39, we know that for −1 < x < 1,

1  ∞
= (−1)n x2n
1 + x2 n=0

Thus for x in this range, we can differentiate both sides, and differentiate the right-hand side term by term, to
get

d 1 −2x  ∞
= = (−1)n 2nx2n−1
dx 1 + x2 (x2 + 1)2 n=1

(Note the change in the lower limit of summation, since the n = 0 term is a constant, whose derivative is
zero.) Cancelling the minus sign on the left gives

2x  ∞
= (−1)n−1 (2n)x2n−1
(x + 1)
2 2
n=1

57. Show that the following series converges absolutely for |x| < 1 and compute its sum:
F(x) = 1 − x − x2 + x3 − x4 − x5 + x6 − x7 − x8 + · · ·
Hint: Write F(x) as a sum of three geometric series with common ratio x3 .
solution To find the radius of convergence, we compute
   
 an+1   ±1 · xn+1 
ρ = lim   = lim  lim |x| = |x|
 = n→∞
n→∞ an  n→∞  ±1 · xn 

The radius of convergence is therefore R = 1, and the series converges absolutely for |x| < 1.
By Exercise 43 of Section 4, any rearrangement of the terms of an absolutely convergent series yields
another absolutely convergent series with the same sum as the original series. Following the hint, we now
rearrange the terms of F(x) as the sum of three geometric series:
  
F(x) = 1 + x3 + x6 + · · · − x + x4 + x7 + · · · − x2 + x5 + x8 + · · ·

∞ 
∞ 
∞
1 x x2 1 − x − x2
= (x3 )n − x(x3 )n − x2 (x3 )n = − − =
n=0 n=0 n=0
1 − x3 1 − x3 1 − x3 1 − x3

58. Show that for |x| < 1,

1 + 2x
= 1 + x − 2x2 + x3 + x4 − 2x5 + x6 + x7 − 2x8 + · · ·
1 + x + x2
Hint: Use the hint from Exercise 57.
 S E C T I O N 10.6 Power Series (LT SECTION 11.6) 131

solution The terms in the series on the right-hand side are either of the form xn or −2xn for some n. Because
n n
lim |2xn | = lim |xn | = |x|
n→∞ n→∞

it follows that
n
lim |an | = |x|
n→∞

Hence, by the Root Test, the series converges absolutely for |x| < 1.
By Exercise 43 of Section 4, any rearrangement of the terms of an absolutely convergent series yields
another absolutely convergent series with the same sum as the original series. If we let S denote the sum of
the series, then
  
S = 1 + x3 + x6 + · · · + x + x4 + x7 + · · · − 2 x2 + x5 + x8 + · · ·

1 x 2x2 1 + x − 2x2 (1 − x)(2x + 1) 2x + 1

= + − = = =
1−x 3 1−x 3 1−x 3 1 − x3 (1 − x)(1 + x + x2 ) 1 + x + x2

∞
xn
2

59. Find all values of x such that converges.

n=1
n!
n2
solution With an = xn! ,
 
 an+1  = |x|
(n+1)2
n! |x|2n+1
 an  (n + 1)! |x|n2 · =
n+1
if |x| ≤ 1, then
|x|2n+1
lim =0
n→∞ n + 1

and the series converges absolutely. On the other hand, if |x| > 1, then

|x|2n+1
lim =∞
n→∞ n + 1

∞
xn
2

and the series diverges. Thus, converges for −1 ≤ x ≤ 1 and diverges elsewhere.
n=1
n!
60. Find all values of x such that the following series converges:

F(x) = 1 + 3x + x2 + 27x3 + x4 + 243x5 + · · ·

solution Observe that F(x) can be written as the sum of two geometric series:

  
∞ 
∞
F(x) = 1 + x2 + x4 + · · · + 3x + 27x3 + 243x5 + · · · = (x2 )n + 3x(9x2 )n
n=0 n=0

The first geometric series converges for |x2 | < 1, or |x| < 1; the second geometric series converges for
|9x2 | < 1, or |x| < 13 . Since both geometric series must converge for F(x) to converge, we find that F(x)
converges for |x| < 13 , the intersection of the intervals of convergence for the two geometric series.

∞
61. Find a power series P(x) = an xn satisfying the differential equation y = −y with initial condition
n=0
y(0) = 1. Then use Eq. (8) in Section 9.1 to conclude that P(x) = e−x .

∞
solution Let P(x) = an xn and note that P(0) = a0 ; thus, to satisfy the initial condition P(0) = 1, we
n=0
must take a0 = 1. Now,

∞
P (x) = nan xn−1
n=1
132 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

so

∞ 
∞ 
∞
P (x) + P(x) = nan xn−1 + an xn = [(n + 1)an+1 + an ] xn
n=1 n=0 n=0
n
In order for this series to be equal to zero, the coefficient of x must be equal to zero for each n; thus
an
(n + 1)an+1 + an = 0 or an+1 = −
n+1
Starting from a0 = 1, we then calculate
a0
a1 = − = −1;
1
a1 1
a2 = − = ;
2 2
a2 1 1
a3 = − = − = −
3 6 3!
and, in general,
1
an = (−1)n
n!
Hence,
∞
xn
P(x) = (−1)n
n=0
n!
The solution to the initial value problem y = −y, y(0) = 1 is y = e−x . Because this solution is unique, it
follows that
∞
xn
P(x) = (−1)n = e−x
n=0
n!

x2 x4 x6
62. Let C(x) = 1 − + − + ··· .
2! 4! 6!
(a) Show that C(x) has an infinite radius of convergence.
(b) Prove that C(x) and f (x) = cos x are both solutions of y = −y with initial conditions y(0) = 1, y (0) = 0.
This Initial Value Problem has a unique solution, so we have C(x) = cos x for all x.
solution
(a) Consider the series
x2 x4 x6 ∞
x2n
C(x) = 1 − + − + ··· = (−1)n
2! 4! 6! n=0
(2n)!
2n
With an = (−1)n (2n)
x
!,
 
 an+1  = |x| |x|2
2n+2
(2n)!
·
 an  (2n + 2)! |x|2n =
(2n + 2)(2n + 1)
and  
an+1 
r = lim  =0
n→∞ an 

The radius of convergence for C(x) is therefore R = r−1 = ∞.

(b) Differentiating the series defining C(x) term-by-term, we find
∞
x2n−1 
∞
x2n−1
C  (x) = (−1)n (2n) = (−1)n
n=1
(2n)! n=1 (2n − 1)!
and
∞
x2n−2 
∞
x2n−2
C  (x) = (−1)n (2n − 1) = (−1)n
n=1
(2n − 1)! n=1 (2n − 2)!
∞
x2n ∞
x2n
= (−1)n+1 = − (−1)n = −C(x)
n=0
(2n)! n=0
(2n)!

Moreover, C(0) = 1 and C  (0) = 0.

 S E C T I O N 10.6 Power Series (LT SECTION 11.6) 133

63. Use the power series for y = e x to show that

1 1 1 1
= − + − ···
e 2! 3! 4!
Use your knowledge of alternating series to find an N such that the partial sum S N approximates e−1 to within
an error of at most 10−3 . Confirm this using a calculator to compute both S N and e−1 .
solution Recall that the series for e x is

∞
xn x2 x3 x4
=1+x+ + + + ···
n=0
n! 2! 3! 4!

Setting x = −1 yields
1 1 1 1 1 1
e−1 = 1 − 1 + − + − +··· = − + − +···
2! 3! 4! 2! 3! 4!
This is an alternating series with an = 1
(n+1)!
. The error in approximating e−1 with the partial sum S N is
therefore bounded by
1
|S N − e−1 | ≤ aN+1 =
(N + 2)!

To make the error at most 10−3 , we must choose N such that

1
≤ 10−3 or (N + 2)! ≥ 1000
(N + 2)!
For N = 4, (N + 2)! = 6! = 720 < 1000, but for N = 5, (N + 2)! = 7! = 5040; hence, N = 5 is the smallest
value that satisfies the Error Bound. The corresponding approximation is
1 1 1 1 1
S5 = − + − + = 0.368055555
2! 3! 4! 5! 6!
Now, e−1 = 0.367879441, so

|S 5 − e−1 | = 1.761 × 10−4 < 10−3


64. Let P(x) = an xn be a power series solution to y = 2xy with initial condition y(0) = 1.
n=0
(a) Show that the odd coefficients a2k+1 are all zero.
(b) Prove that a2k = a2k−2 /k and use this result to determine the coefficients a2k .

∞
solution Let P(x) = an xn and note that P(0) = a0 ; thus, to satisfy the initial condition P(0) = 1, we
n=0
must take a0 = 1. Now,

∞
P (x) = nan xn−1
n=1

so

∞ 
∞ 
∞ 
∞
P (x) − 2xP(x) = nan xn−1 − 2an xn+1 = nan xn−1 − 2an−2 xn−1
n=1 n=0 n=1 n=2


∞
= a1 + [nan − 2an−2 ] xn−1
n=2

In order for this series to be equal to zero, the coefficient of xn must be equal to zero for each n; thus, a1 = 0
and
2an−2
nan − 2an−2 = 0 or an =
n
134 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

(a) We know that a1 = 0 and

2an−2
an =
n
Thus,
2a1
a3 = = 0;
3
2a3
a5 = = 0;
5
2a5
a7 = =0
7
and, in general, a2k+1 = 0 for all k.
(b) Replace n by 2k in the equation
2an−2 2a2k−2 a2k−2
an = to obtain a2k = =
n 2k k
Starting from a0 = 1, we then calculate
a0 1
a2 = =1= ;
1 1!
a2 1 1
a4 = = = ;
2 2 2!
a4 1 1
a6 = = =
3 6 3!
and, in general, a2k = k1! .
65. Find a power series P(x) satisfying the differential equation
y − xy + y = 0 10
with initial condition y(0) = 1, y (0) = 0. What is the radius of convergence of the power series?
∞
solution Let P(x) = an xn . Then
n=0


∞ 
∞
P (x) = nan xn−1 and P (x) = n(n − 1)an xn−2
n=1 n=2

Note that P(0) = a0 and P (0) = a1 ; in order to satisfy the initial conditions P(0) = 1, P (0) = 0, we must
have a0 = 1 and a1 = 0. Now,

∞ 
∞ 
∞
P (x) − xP (x) + P(x) = n(n − 1)an xn−2 − nan xn + an xn
n=2 n=1 n=0


∞ 
∞ 
∞
= (n + 2)(n + 1)an+2 xn − nan xn + an xn
n=0 n=1 n=0


∞
= 2a2 + a0 + [(n + 2)(n + 1)an+2 − nan + an ] xn
n=1

In order for this series to be equal to zero, the coefficient of xn must be equal to zero for each n; thus,
2a2 + a0 = 0 and (n + 2)(n + 1)an+2 − (n − 1)an = 0, or
1 n−1
a2 = − a0 and an+2 = an
2 (n + 2)(n + 1)
Since a1 = 0 and an+2 is a multiple of an for all n, we conclude that an = 0 if n is odd. As for the even
coefficients, we have a0 = 1, a2 = − 12 ,
1 1 3 3 5 15
a4 = a2 = − ; a6 = a4 = − ; a8 = a6 = −
(4)(3) 4! (6)(5) 6! (8)(7) 8!
 S E C T I O N 10.6 Power Series (LT SECTION 11.6) 135

and so on. Claim that in general

1 · 3 · · · (2n − 3)
a2n = − for n ≥ 2
(2n)!
We prove this by induction on n. First, the computation above shows that it holds for n = 2, n = 3, and n = 4.
Now assume it holds for n = k, and compute:

2k − 1 2k − 1 1 · 3 · · · (2k − 3)
a2(k+1) = a2k+2 = a2k = −
(2k + 2)(2k + 1) (2k + 2)(2k + 1) (2k)!
1 · 3 · · · (2k − 1) 1 · 3 · · · (2(k + 1) − 1)
=− =−
(2k + 2)! (2(k + 1))!
and the proof is complete. Thus

1 2  1 · 3 · · · (2n − 3) 2n
∞
P(x) = 1 − x − x
2 n=2
(2n)!

To compute the radius of convergence, since we are interested in the behavior as n → ∞, we may assume
n ≥ 4; then treating the series as a power series in x2 we get
   
 a2n+2   1 · 3 · · · (2n − 3) · (2n − 1)x2n+2 (2n)! 
ρ = lim   = lim  · 
n→∞ a2n n→∞  (2(n + 1))! 1 · 3 · · · (2n − 3)x 
2n
 
 (2n − 1)x2 
= lim  
n→∞  (2n + 1)(2n + 2) 

2n − 1
= |x|2 lim =0
n→∞ 4n2 + 6n + 2
Thus, the radius of convergence is R = ∞.
66. Find a power series satisfying Eq. (10) with initial condition y(0) = 0, y (0) = 1.
∞
solution Let P(x) = an xn be a solution to Eq. (10). Similar to the previous exercise, the initial conditions
n=0
P(0) = 0, P (0) = 1 imply that a0 = 0 and a1 = 1. Also from the previous exercise, we know that
1 n−1
a2 = − a0 and an+2 = an
2 (n + 2)(n + 1)
Since a0 = 0 and an+2 is a multiple of an for all n, it follows that all the even coefficients are zero. Also,
1−1
a3 = a1 = 0
3·2
again since an+2 is a multiple of an for all n, it follows that all odd coefficients for n ≥ 3 are also zero. The
only remaining nonzero coefficient is a1 = 1, so that

P(x) = x

67. Prove that


∞
(−1)k
J2 (x) = x2k+2
k=0
22k+2 k! (k + 3)!

is a solution of the Bessel differential equation of order 2:

x2 y + xy + (x2 − 4)y = 0


∞
(−1)k
solution Let J2 (x) = x2k+2 . Then
k=0
22k+2 k! (k + 2)!

∞
(−1)k (k + 1) 2k+1
J2 (x) = x
k=0
22k+1 k! (k + 2)!
136 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)


∞
(−1)k (k + 1)(2k + 1)
J2 (x) = x2k
k=0
22k+1 k! (k + 2)!

and

∞
(−1)k (k + 1)(2k + 1) ∞
(−1)k (k + 1) 2k+2
x2 J2 (x) + xJ2 (x) + (x − 4)J2 (x) =
2
x 2k+2
+ x
k=0
22k+1 k! (k + 2)! k=0
22k+1 k! (k + 2)!

∞
(−1)k  ∞
(−1)k
− x2k+4 − x2k+2
k=0
22k+2 k! (k + 2)! k=0
22k k! (k + 2)!

∞
(−1)k k(k + 2) 2k+2 
∞
(−1)k−1
= x + x2k+2
k=0
2 k!(k + 2)!
2k
k=1
22k (k − 1)! (k + 1)!

∞
(−1)k  ∞
(−1)k
= x2k+2 − x2k+2 = 0
k=1
2 (k − 1)!(k + 1)!
2k
k=1
22k (k − 1)!(k + 1)!

68. Why is it impossible to expand f (x) = |x| as a power series that converges in an interval around
x = 0? Explain using Theorem 2.
solution Suppose that there exists a c > 0 such that f can be represented by a power series on the interval
(−c, c); that is,

∞
|x| = an xn
n=0

for |x| < c. Then it follows by Theorem 2 that |x| is differentiable on (−c, c). This contradicts the well-known
property that f (x) = |x| is not differentiable at the point x = 0.

Further Insights and Challenges


∞
69. Suppose that the coefficients of F(x) = an xn are periodic; that is, for some whole number M > 0, we
n=0
have a M+n = an . Prove that F(x) converges absolutely for |x| < 1 and that
a0 + a1 x + · · · + a M−1 x M−1
F(x) =
1 − xM
Hint: Use the hint for Exercise 57.
solution Suppose the coefficients of F(x) are periodic, with a M+n = an for some whole number M and all
n. The F(x) can be written as the sum of M geometric series:
 
F(x) = a0 1 + x M + x2M + · · · + a1 x + x M+1 + x2M+1 + · · · +
 
= a2 x2 + x M+2 + x2M+2 + · · · + · · · + a M−1 x M−1 + x2M−1 + x3M−1 + · · ·

a0 a1 x a2 x2 a M−1 x M−1 a0 + a1 x + a2 x2 + · · · + a M−1 x M−1

= + + + · · · + =
1 − xM 1 − xM 1 − xM 1 − xM 1 − xM
As each geometric series converges absolutely for |x| < 1, it follows that F(x) also converges absolutely for
|x| < 1.
∞
70. Continuity of Power Series Let F(x) = an xn be a power series with radius of convergence R > 0.
(a) Prove the inequality n=0

|xn − yn | ≤ n|x − y|(|x|n−1 + |y|n−1 ) 11

Hint: xn − yn = (x − y)(xn−1 + xn−2 y + · · · + yn−1 ).

∞
(b) Choose R1 with 0 < R1 < R. Show that the infinite series M = 2n|an |Rn1 converges. Hint: Show that
n=0
n|an |Rn1 < |an |xn for all n sufficiently large if R1 < x < R.
(c) Use the inequality in (11) to show that if |x| < R1 and |y| < R1 , then |F(x) − F(y)| ≤ M|x − y|.
 S E C T I O N 10.6 Power Series (LT SECTION 11.6) 137

(d) Prove that if |x| < R, then F is continuous at x. Hint: Choose R1 such that |x| < R1 < R. Show that if
 > 0 is given, then |F(x) − F(y)| ≤  for all y such that |x − y| < δ, where δ is any positive number that is
less than /M and R1 − |x| (see Figure 6).

( ( ) ) x
−R 0 x R1 R

FIGURE 6 If x > 0, choose δ > 0 less than /M and R1 − x.

solution
(a) Take the absolute value of both sides of the identity

xn − yn = (x − y)(xn−1 + xn−2 y + · · · + yn−1 )

and then apply the triangle inequality to obtain


|xn − yn | ≤ |x − y| |x|n−1 + |x|n−2 |y| + |x|n−3 |y|2 + · · · + |x||y|n−2 + |y|n−1

Now, if |x| ≤ |y| then |x|n−k |y|k−1 ≤ |y|n−k |y|k−1 = |y|n−1 , and if |y| ≤ |x| then |x|n−k |y|k−1 ≤ |x|n−k |x|k−1 = |x|n−1 . In
either case, |x|n−k |y|k−1 ≤ |x|n−1 + |y|n−1 . Thus,

|xn − yn | ≤ |x − y| |x|n−1 + (n − 2)(|x|n−1 + |y|n−1 ) + |y|n−1
 
= (n − 1)|x − y| |x|n−1 + |y|n−1 ≤ n|x − y| |x|n−1 + |y|n−1

(b) Let 0 < R1 < R. Then,

 
2(n + 1)|an+1 |Rn+1 n + 1  an+1 
ρ = lim 1
= R 1 lim ·  an 

n→∞ 2n|an |Rn1 n→∞ n
 
Since the radius of convergence is R, we must have lim  aan+1
n
 = 1 , so that the limit is
R
n→∞
 
n + 1  an+1  1 R
R1 lim ·   = R1 · 1 · = 1 < 1
n→∞ n an  R R

∞
Thus, the series M = 2n|an |Rn1 converges by the Ratio Test.
n=0
(c) Suppose |x| < R1 and |y| < R1 . Then
 
 ∞ ∞  ∞ ∞ 
|F(x) − F(y)| =  an x − n
an y  ≤
n
|an ||xn − yn | ≤ n|an ||x − y| |x|n−1 + |y|n−1
 n=0 n=0
 n=0 n=0


∞ 
≤ |x − y| n|an | Rn−1
1 + Rn−1
1 = M|x − y|
n=0


∞
(d) Let |x| < R, and let R1 be a number such that |x| < R1 < R. Then by part (b), M = 2n|an |Rn1 is finite
n=0
and by part (c)

|F (x) − F (y) | ≤ M|x − y|


for |y| < R1 . Now, let  > 0, and choose δ > 0 so that δ < and δ < R1 − |x|. Then, whenever |y − x| < δ,
M
|y| = | (y − x) + x| ≤ |y − x| + |x| < δ + |x| < R1

so

|F (x) − F (y) | < M|x − y| < Mδ < M · =
M
Thus, F is continuous at x.
138 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

10.7 Taylor Polynomials (LT Section 11.7)

Preliminary Questions
1. What is T 3 centered at a = 3 for a function f such that f (3) = 9, f  (3) = 8, f  (3) = 4, and f  (3) = 12?
solution In general, with a = 3,
f  (3) f  (3)
T 3 (x) = f (3) + f  (3)(x − 3) + (x − 3)2 + (x − 3)3
2 6
Using the information provided, we find
T 3 (x) = 9 + 8(x − 3) + 2(x − 3)2 + 2(x − 3)3

2. The dashed graphs in Figure 7 are Taylor polynomials for a function f . Which of the two is a Maclaurin
polynomial?
y y

y = f (x) y = f (x)
−1 1 2 3 −1 1 3
x x
2

(A) (B)
FIGURE 7

solution A Maclaurin polynomial always gives the value of f (0) exactly. This is true for the Taylor poly-
nomial sketched in (B); thus, this is the Maclaurin polynomial.
3. For which value of x does the Maclaurin polynomial T n satisfy T n (x) = f (x), no matter what f is?
solution A Maclaurin polynomial always gives the value of f (0) exactly.
4. Let T n be the Maclaurin polynomial of a function f satisfying | f (4) (x)| ≤ 1 for all x. Which of the
following statements follow from the Error Bound?
2
(a) |T 4 (2) − f (2)| ≤
3
2
(b) |T 3 (2) − f (2)| ≤
3
1
(c) |T 3 (2) − f (2)| ≤
3
solution For a function f (x) satisfying | f (4) (x)| ≤ 1 for all x,
1 (4) 16 2
|T 3 (2) − f (2)| ≤ | f (x)|24 ≤ <
24 24 3
Thus, (b) is the correct answer.

Exercises
In Exercises 1–16, calculate the Taylor polynomials T 2 and T 3 centered at x = a for the given function and
value of a.

1. f (x) = sin x, a=0

solution First, we calculate and evaluate the needed derivatives:

f (x) = sin x f (a) = 0
f  (x) = cos x f  (a) = 1
f  (x) = − sin x f  (a) = 0
f  (x) = − cos x f  (a) = −1
 S E C T I O N 10.7 Taylor Polynomials (LT SECTION 11.7) 139

Now,

f  (a) 0
T 2 (x) = f (a) + f  (a)(x − a) + (x − a)2 = 0 + 1(x − 0) + (x − 0)2 = x; and
2 2
 
f (a) f (a)
T 3 (x) = f (a) + f  (a)(x − a) + (x − a)2 + (x − a)3
2 6
0 −1 1
= 0 + 1(x − 0) + (x − 0)2 + (x − 0)3 = x − x3
2 6 6

π
2. f (x) = sin x, a=
2
solution First, we calculate and evaluate the needed derivatives:

f (x) = sin x f (a) = 1

f  (x) = cos x f  (a) = 0
f  (x) = − sin x f  (a) = −1
f  (x) = − cos x f  (a) = 0

Now,

f  (a)
T 2 (x) = f (a) + f  (a)(x − a) + (x − a)2
2
 π  −1  π 2 1 π 2
=1+0 x− + x− =1− x− ; and
2 2 2 2 2
f  (a) f  (a)
T 3 (x) = f (a) + f  (a)(x − a) + (x − a)2 + (x − a)3
2 6
 π  −1  π 2 0  π 3 1 π 2
=1+0 x− + x− + x− =1− x−
2 2 2 6 2 2 2

1
3. f (x) = , a=2
1+x
solution First, we calculate and evaluate the needed derivatives:

1 1
f (x) = f (a) =
1+x 3
−1 1
f  (x) = f  (a) = −
(1 + x)2 9
2 2
f  (x) = f  (a) =
(1 + x)3 27
−6 2
f  (x) = f  (a) = −
(1 + x)4 27

Now,

f  (a) 1 1 2/27
T 2 (x) = f (a) + f  (a)(x − a) + (x − a)2 = − (x − 2) + (x − 2)2
2! 3 9 2!
1 1 1
= − (x − 2) + (x − 2)2
3 9 27
f  (a) f  (a)
T 3 (x) = f (a) + f  (a)(x − a) + (x − a)2 + (x − a)3
2! 3!
1 1 2/27 2/27 1 1 1 1
= − (x − 2) + (x − 2)2 − (x − 2)3 = − (x − 2) + (x − 2)2 − (x − 2)3
3 9 2! 3! 3 9 27 81
140 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

1
4. f (x) = , a = −1
1 + x2
solution First, we calculate and evaluate the needed derivatives:
1
f (x) = f (a) = 1/2
1 + x2
−2x
f  (x) = 2 f  (a) = 1/2
(x + 1)2
2(3x2 − 1)
f  (x) = f  (a) = 1/2
(x2 + 1)3
−24x(x2 − 1)
f  (x) = f  (a) = 0
(x2 + 1)4
Now,
f  (a)
T 2 (x) = f (a) + f  (a)(x − a) + (x − a)2
2
1 1 1/2 1 1 1
= + (x + 1) + (x + 1)2 = + (x + 1) + (x + 1)2 ; and
2 2 2 2 2 4
 
f (a) f (a)
T 3 (x) = f (a) + f  (a)(x − a) + (x − a)2 + (x − a)3
2 6
1 1 1/2 0 1 1 1
= + (x + 1) + (x + 1)2 + (x + 1)3 = + (x + 1) + (x + 1)2
2 2 2 6 2 2 4
5. f (x) = x4 − 2x, a=3

solution First calculate and evaluate the needed derivatives:

f (x) = x4 − 2x f (a) = 75

f (x) = 4x − 2
3
f  (a) = 106
f  (x) = 12x2 f  (a) = 108
f  (x) = 24x f  (a) = 72
Now,
f  (a) 108
T 2 (x) = f (a) + f  (a)(x − a) + (x − a)2 = 75 + 106(x − 3) + (x − 3)2
2 2
= 75 + 106(x − 3) + 54(x − 3)2
f  (a) f  (a)
T 3 (x) = f (a) + f  (a)(x − a) + (x − a)2 + (x − a)3
2 3!
108 72
= 75 + 106(x − 3) + (x − 3)2 + (x − 3)3
2 3!
= 75 + 106(x − 3) + 54(x − 3)2 + 12(x − 3)3

x2 + 1
6. f (x) = , a = −2
x+1
solution First calculate and evaluate the needed derivatives:
x2 + 1
f (x) = f (a) = −5
x+1
x2 + 2x − 1
f  (x) = f  (a) = −1
(x + 1)2
4
f  (x) = f  (a) = −4
(x + 1)3
−12
f  (x) = f  (a) = −12
(x + 1)4
 S E C T I O N 10.7 Taylor Polynomials (LT SECTION 11.7) 141

Now,
f  (a) −4
T 2 (x) = f (a) + f  (a)(x − a) + (x − a)2 = −5 − (x + 2) + (x + 2)2
2! 2
= −5 − (x + 2) − 2(x + 2)2
f  (a) f  (a)
T 3 (x) = f (a) + f  (a)(x − a) + (x − a)2 + (x − a)3
2! 3!
−4 −12
= −5 − (x + 2) + (x + 2)2 + (x + 2)3
2 3!
= −5 − (x + 2) − 2(x + 2)2 − 2(x + 2)3

√
7. f (x) = x, a=1

solution First calculate and evaluate the needed derivatives:

√
f (x) = x f (a) = 1
1 −1/2 1
f  (x) = x f  (a) =
2 2
1 1
f  (x) = − x−3/2 f  (a) = −
4 4
3 3
f  (x) = x−5/2 f  (a) =
8 8
Now,
f  (a) 1 −1/4
T 2 (x) = f (a) + f  (a)(x − a) + (x − a)2 = 1 + (x − 1) + (x − 1)2
2! 2 2
1 1
= 1 + (x − 1) − (x − 1)2
2 8
f  (a) f  (a)
T 3 (x) = f (a) + f  (a)(x − a) + (x − a)2 + (x − a)3
2! 3!
1 −1/4 3/8
= 1 + (x − 1) + (x − 1)2 + (x − 1)3
2 2 3!
1 1 1
= 1 + (x − 1) − (x − 1)2 + (x − 1)3
2 8 16

√
8. f (x) = x, a=9

solution First calculate and evaluate the needed derivatives:

√
f (x) = x f (a) = 3
1 −1/2 1
f  (x) = x f  (a) =
2 6
1 1
f  (x) = − x−3/2 f  (a) = −
4 108
3 1
f  (x) = x−5/2 f  (a) =
8 648
Now,
f  (a) 1 −1/108
T 2 (x) = f (a) + f  (a)(x − a) + (x − a)2 = 3 + (x − 9) + (x − 9)2
2! 6 2
1 1
= 3 + (x − 9) − (x − 9)2
6 216
f  (a) f  (a)
T 3 (x) = f (a) + f  (a)(x − a) + (x − a)2 + (x − a)3
2! 3!
142 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

1 −1/108 1/648
= 3 + (x − 9) + (x − 9)2 + (x − 9)3
6 2 3!
1 1 1
= 3 + (x − 9) − (x − 9)2 + (x − 9)3
6 216 3888

9. f (x) = tan x, a=0

solution First, we calculate and evaluate the needed derivatives:

f (x) = tan x f (a) = 0

f (x) = sec x 2
f  (a) = 1
f  (x) = 2 sec2 x tan x f  (a) = 0
f  (x) = 2 sec4 x + 4 sec2 x tan2 x f  (a) = 2
Now,
f  (a) 0
T 2 (x) = f (a) + f  (a)(x − a) + (x − a)2 = 0 + 1(x − 0) + (x − 0)2 = x; and
2 2
 
f (a) f (a)
T 3 (x) = f (a) + f  (a)(x − a) + (x − a)2 + (x − a)3
2 6
0 2 1
= 0 + 1(x − 0) + (x − 0)2 + (x − 0)3 = x + x3
2 6 3

π
10. f (x) = tan x, a=
4
solution First, we calculate and evaluate the needed derivatives:
f (x) = tan x f (a) = 1
f  (x) = sec2 x f  (a) = 2
f  (x) = 2 sec2 x tan x f  (a) = 4
f  (x) = 2 sec4 x + 4 sec2 x tan2 x f  (a) = 16
Now,
f  (a)  π 4  π 2
T 2 (x) = f (a) + f  (a)(x − a) + (x − a)2 = 1 + 2 x − + x−
2 4 2 4
 π   π  2
=1+2 x− +2 x− ; and
4 4
f  (a) f  (a)
T 3 (x) = f (a) + f  (a)(x − a) + (x − a)2 + (x − a)3
2 6
 π 4  π 2 16  π 3  π  π 2 8  π 3
=1+2 x− + x− + x− =1+2 x− +2 x− + x−
4 2 4 6 4 4 4 3 4

11. f (x) = e−x + e−2x , a=0

solution First, we calculate and evaluate the needed derivatives:

f (x) = e−x + e−2x f (a) = 2

 −x −2x
f (x) = −e − 2e f  (a) = −3
f  (x) = e−x + 4e−2x f  (a) = 5
f  (x) = −e−x − 8e−2x f  (a) = −9
Now,
f  (a)
T 2 (x) = f (a) + f  (a)(x − a) + (x − a)2
2
 S E C T I O N 10.7 Taylor Polynomials (LT SECTION 11.7) 143

5 5
= 2 + (−3)(x − 0) + (x − 0)2 = 2 − 3x + x2 ; and
2 2

f (a) f  (a)
T 3 (x) = f (a) + f  (a)(x − a) + (x − a)2 + (x − a)3
2 6
5 −9 5 3
= 2 + (−3)(x − 0) + (x − 0)2 + (x − 0)3 = 2 − 3x + x2 − x3
2 6 2 2

12. f (x) = e2x , a = ln 2

solution First calculate and evaluate the needed derivatives:

f (x) = e2x f (a) = 4

f  (x) = 2e2x f  (a) = 8
f  (x) = 4e2x f  (a) = 16
f  (x) = 8e2x f  (a) = 32
Now
f  (a) 16
T 2 (x) = f (a) + f  (a)(x − a) + (x − a)2 = 4 + 8(x − ln 2) + (x − ln 2)2
2! 2!
= 4 + 8(x − ln 2) + 8(x − ln 2)2
f  (a) f  (a)
T 3 (x) = f (a) + f  (a)(x − a) + (x − a)2 + (x − a)3
2! 3!
16 32
= 4 + 8(x − ln 2) + (x − ln 2)2 + (x − ln 2)3
2! 6
16
= 4 + 8(x − ln 2) + 8(x − ln 2)2 + (x − ln 2)3
3

13. f (x) = x2 e−x , a = 1

solution First, we calculate and evaluate the needed derivatives:

f (x) = x2 e−x f (a) = 1/e

f  (x) = (2x − x2 )e−x f  (a) = 1/e
f  (x) = (x2 − 4x + 2)e−x f  (a) = −1/e
f  (x) = (−x2 + 6x − 6)e−x f  (a) = −1/e
Now,
f  (a)
T 2 (x) = f (a) + f  (a)(x − a) + (x − a)2
2
1 1 −1/e 1 1 1
= + (x − 1) + (x − 1)2 = + (x − 1) − (x − 1)2 ; and
e e 2 e e 2e
f  (a) f  (a)
T 3 (x) = f (a) + f  (a)(x − a) + (x − a)2 + (x − a)3
2 6
 
1 1 −1/e −1/e
= + (x − 1) + (x − 1)2 + (x − 1)3
e e 2 6
1 1 1 1
= + (x − 1) − (x − 1)2 − (x − 1)3
e e 2e 6e

14. f (x) = cosh 2x, a=0

solution First calculate and evaluate the needed derivatives:

f (x) = cosh 2x f (a) = 1
144 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

f  (x) = 2 sinh 2x f  (a) = 0

f  (x) = 4 cosh 2x f  (a) = 4
f  (x) = 8 sinh 2x f  (a) = 0
so that
f  (a) 4
T 2 (x) = f (a) + f  (a)(x − a) + (x − a)2 = 1 + 0(x − 0) + (x − 0)2 = 1 + 2x2 ; and
2! 2!
 
f (a) f (a)
T 3 (x) = f (a) + f  (a)(x − a) + (x − a)2 + (x − a)3
2! 3!
0
= 1 + 0(x − 0) + 2(x − 0)2 + (x − 0)3 = 1 + 2x2
3!

ln x
15. f (x) = , a=1
x
solution First calculate and evaluate the needed derivatives:
ln x
f (x) = f (a) = 0
x
1 − ln x
f  (x) = f (a) = 1
x2
−3 + 2 ln x
f  (x) = f (a) = −3
x3
11 − 6 ln x
f  (x) = f (a) = 11
x4
so that
f  (a) −3
T 2 (x) = f (a) + f  (a)(x − a) + (x − a)2 = 0 + 1(x − 1) + (x − 1)2
2! 2!
3
= (x − 1) − (x − 1)2 ; and
2
f  (a) f  (a)
T 3 (x) = f (a) + f  (a)(x − a) + (x − a)2 + (x − a)3
2! 3!
−3 11
= 0 + 1(x − 1) + (x − 1)2 + (x − 1)3
2! 3!
3 11
= (x − 1) − (x − 1)2 + (x − 1)3
2 6
16. f (x) = ln(x + 1), a=0

solution First, we calculate and evaluate the needed derivatives:

f (x) = ln(x + 1) f (a) = 0
1
f  (x) = f  (a) = 1
x+1
−1
f  (x) = f  (a) = −1
(x + 1)2
2
f  (x) = f  (a) = 2
(x + 1)3
Now,
f  (a) −1 1
T 2 (x) = f (a) + f  (a)(x − a) + (x − a)2 = 0 + 1(x − 0) + (x − 0)2 = x − x2 ; and
2 2 2
 
f (a) f (a)
T 3 (x) = f (a) + f  (a)(x − a) + (x − a)2 + (x − a)3
2 6
−1 2 1 1
= 0 + 1(x − 0) + (x − 0)2 + (x − 0)3 = x − x2 + x3
2 6 2 3
 S E C T I O N 10.7 Taylor Polynomials (LT SECTION 11.7) 145

17. Show that the second Taylor polynomial for f (x) = px2 + qx + r, centered at a = 1, is f (x).
solution Calculating the derivatives at a = 1, we get

f (x) = px2 + qx + r f (1) = p + q + r


f (x) = 2px + q f  (1) = 2p + q
f  (x) = 2p f  (1) = 2p
Now,
f  (a)
T 2 (x) = f (a) + f  (a)(x − a) + (x − a)2
2
2p
= p + q + r + (2p + q)(x − 1) + (x − 1)2
2
= px2 + qx + r = f (x)

18. Show that the third Maclaurin polynomial for f (x) = (x − 3)3 is f (x).
solution Calculating the derivatives at a = 0, we get

f (x) = (x − 3)3 f (0) = −27

f  (x) = 3(x − 3)2 f  (0) = 27
f  (x) = 6(x − 3) f  (0) = −18
f  (x) = 6 f  (0) = 6
Now,
f  (a) f  (a)
T 2 (x) = f (a) + f  (a)(x − a) + (x − a)2 + (x − a)3
2 6
18 2 6 3
= −27 + 27x − x + x
2 6
= x3 − 9x2 + 27x − 27 = f (x)

19. Show that the nth Maclaurin polynomial for f (x) = e x is

x x2 xn
T n (x) = 1 + + + ··· +
1! 2! n!
solution With f (x) = e x , it follows that f (n) (x) = e x and f (n) (0) = 1 for all n. Thus,

1 1 x2 xn
T n (x) = 1 + 1(x − 0) + (x − 0)2 + · · · + (x − 0)n = 1 + x + + ··· +
2 n! 2 n!

1
20. Show that the nth Taylor polynomial for f (x) = at
x+1
a = 1 is
1 (x − 1) (x − 1)2 (x − 1)n
T n (x) = − + + · · · + (−1)n n+1
2 4 8 2
solution Let f (x) = 1
x+1 . Then

1 1 (−1)0 0!
f (x) = f (1) = =
x+1 2 20+1
−1 1 (−1)1 1!
f  (x) = f  (1) = − =
(x + 1)2 4 21+1
2 1 (−1)2 2!
f  (x) = f  (1) = =
(x + 1)3 4 22+1
.. ..
. .
146 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

(−1)n n! (−1)n n!
f (n) (x) = f (n) (1) =
(x + 1)n+1 2n+1
Therefore,
 
1 1 1 (x − 1)2 (−1)n n! (x − 1)n
T n (x) = + − (x − 1) + + ··· +
2 4 4 2! 2n+1 n!
1 1 (x − 1)2 (x − 1)n
= − (x − 1) + + · · · + (−1)n n+1
2 4 8 2

21. Show that the Maclaurin polynomials for f (x) = sin x are

x3 x5 x2n+1
T 2n+1 (x) = T 2n+2 (x) = x − + − · · · + (−1)n
3! 5! (2n + 1)!
solution Let f (x) = sin x. Then

f (x) = sin x f (0) = 0


f (x) = cos x f  (0) = 1
f  (x) = − sin x f  (0) = 0
f  (x) = − cos x f  (0) = −1
f (4) (x) = sin x f (4) (0) = 0
f (5) (x) = cos x f (5) (0) = 1
.. ..
. .

Consequently,

x3 x5 x2n−1
T 2n−1 (x) = x − + + · · · + (−1)n+1
3! 5! (2n − 1)!
and
x3 x5 x2n−1
T 2n (x) = x − + + · · · + (−1)n+1 + 0 = T 2n−1 (x)
3! 5! (2n − 1)!

22. Show that the Maclaurin polynomials for f (x) = ln(1 + x) are

x2 x3 xn
T n (x) = x − + + · · · + (−1)n−1
2 3 n
solution Let f (x) = ln(1 + x). Then

f (x) = ln(1 + x) f (0) = 0

f  (x) = (1 + x)−1 f  (0) = 1
f  (x) = −(1 + x)−2 f  (0) = −1
f  (x) = 2(1 + x)−3 f  (0) = 2
f (4) (x) = −3!(1 + x)−4 f (4) (0) = −6
f (5) (x) = 4!(1 + x)−5 f (5) (0) = 24

so that in general

f (n) (x) = (−1)n−1 (n − 1)!(1 + x)−n f (n) (0) = (−1)n−1 (n − 1)!

Thus
1 2 2 3 (−1)n−1 (n − 1)! n x2 x3 xn
T n (x) = x − x + x − ··· + x = x− + + · · · + (−1)n−1
2! 3! n! 2 3 n
 S E C T I O N 10.7 Taylor Polynomials (LT SECTION 11.7) 147

In Exercises 23–30, find T n centered at x = a for all n.

1
23. f (x) = , a=0
1+x
solution We have

1
= (ln(1 + x))
1+x
so that from Exercise 22, letting g(x) = ln(1 + x),

f (n) (x) = g(n+1) (x) = (−1)n n!(x + 1)−1−n and f (n) (0) = (−1)n n!

Then

f  (0) 2 f (n) (0) n

T n (x) = f (0) + f  (0)x + x + ··· + x
2! n!
2! 3! n!
= 1 − x + x2 − x3 + · · · + (−1)n xn
2! 3! n!
= 1 − x + x2 − x3 + · · · + (−1)n xn

1
24. f (x) = , a=4
x−1
solution Let f (x) = 1
x−1 . Then

1 1 (−1)0 0!
f (x) = f (4) = =
x−1 3 30+1
−1 1 (−1)1 1!
f  (x) = f  (4) = − =
(x − 1)2 9 31+1
2 2 (−1)2 2!
f  (x) = f  (4) = =
(x − 1)3 27 32+1
.. ..
. .
(−1)n n! (−1)n n!
f (n) (x) = f (n) (4) =
(x − 1)n+1 3n+1

Therefore,
 
1 1 2/27 (−1)n n! (x − 4)n
T n (x) = + − (x − 4) + (x − 4)2 + · · · +
3 9 2 3n+1 n!
1 1 1 (−1)n
= − (x − 4) + (x − 4)2 + · · · + n+1 (x − 4)n
3 9 27 3

25. f (x) = e x , a=1

solution Let f (x) = e x . Then f (n) (x) = e x and f (n) (1) = e for all n. Therefore,

e e
T n (x) = e + e(x − 1) + (x − 1)2 + · · · + (x − 1)n
2! n!

26. f (x) = e x , a = −2

solution Let f (x) = e x . Then f (n) (x) = e x and f (n) (−2) = e−2 for all n. Therefore,

e−2 e−2
T n (x) = e−2 + e−2 (x + 2) + (x + 2)2 + · · · + (x + 2)n
2! n!
148 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

27. f (x) = x−2 , a = 1

solution Taking derivatives, we get
f (x) = x−2 f (1) = 1
 −3
f (x) = −2x f  (1) = −2
f  (x) = 6x−4 f  (1) = 6
f  (x) = −24x−5 f  (1) = −24
.. ..
. .
f (n) (x) = (−1)n n!x−n−2 f (n) (1) = (−1)n n!
Therefore,
3! 4! (n + 1)!
T n (x) = 1 − 2(x − 1) + (x − 1)2 − (x − 1)3 + · · · + (−1)n (x − 1)n
2! 3! n!
= 1 − 2(x − 1) + 3(x − 1)2 − 4(x − 1)3 + · · · + (−1)n (n + 1)(x − 1)n

28. f (x) = x−2 , a = 2

solution We have
1
f (x) = x−2 f (2) =
4
1
f  (x) = −2x−3 f  (2) = −
4
3
f  (x) = 6x−4 f  (2) =
8
3
f  (x) = −24x−5 f  (2) = −
4
.. ..
. .
(n + 1)!
f (n) (x) = (−1)n (n + 1)!x−n−2 f (n) (2) = (−1)n
2n+2
so that
f  (2) f (n) (2)
T n (x) = f (2) + f  (2)(x − 2) +
(x − 2)2 + · · · + (x − 2)n
2! n!
1 1 3 n+1
= − (x − 2) + (x − 2)2 + · · · + (−1)n n+2 (x − 2)n
4 4 16 2
π
29. f (x) = cos x, a=
4
solution Let f (x) = cos x. Then
1
f (x) = cos x f (π/4) = √
2
1
f  (x) = − sin x f  (π/4) = − √
2
1
f  (x) = − cos x f  (π/4) = − √
2
1
f  (x) = sin x f  (π/4) = √
2
This pattern of four values repeats indefinitely. Thus,
⎧
⎪
⎪
⎪ 1
⎪
⎪ (−1)(n+1)/2 √ , n odd
⎪
⎪
⎨ 2
f (n) (π/4) = ⎪
⎪
⎪
⎪
⎪ 1
⎪
⎩(−1) √ ,
⎪ n/2
n even
2
 S E C T I O N 10.7 Taylor Polynomials (LT SECTION 11.7) 149

and
1 1  π 1  π 2 1  π 3
T n (x) = √ − √ x − − √ x− + √ x− ···
2 2 4 2 2 4 6 2 4
In general, the coefficient of (x − π/4)n is
1
± √
( 2)n!
with the pattern of signs +, −, −, +, +, −, −, . . . .
30. f (θ) = sin 3θ, a=0
solution We have

f (θ) = sin 3θ f (0) = 0


f (θ) = 3 cos 3θ f  (0) = 3
f  (θ) = −9 sin 3θ f  (0) = 0
f  (θ) = −27 cos 3θ f  (0) = −27
f (4) (θ) = 81 sin 3θ f (4) (0) = 0

and in general

f (2n) (θ) = (−1)n 32n sin 3θ f (2n) (0) = 0

f (2n+1) (θ) = (−1)n 32n+1 cos 3θ f (2n+1) (0) = (−1)n 32n+1

Thus
27 3 243 5
T n (x) = 3θ − θ + θ − ...
3! 5!
2n+1
where the coefficient of θ2n+1 is (−1)n (2n+1)!
3
.

In Exercises 31–34, find T 2 and use a calculator to compute the error | f (x) − T 2 (x)| for the given values of a
and x.

31. y = e x , a = 0, x = −0.5
solution Let f (x) = e x . Then f  (x) = e x , f  (x) = e x , f (a) = 1, f  (a) = 1 and f  (a) = 1. Therefore
1 1
T 2 (x) = 1 + 1(x − 0) + (x − 0)2 = 1 + x + x2
2 2
and
1
T 2 (−0.5) = 1 + (−0.5) + (−0.5)2 = 0.625
2
Using a calculator, we find
1
f (−0.5) = √ = 0.606531
e
so

|T 2 (−0.5) − f (−0.5)| = 0.0185

π
32. y = cos x, a = 0, x=
12
solution Let f (x) = cos x. Then f  (x) = − sin x, f  (x) = − cos x, f (a) = 1, f  (a) = 0, and f  (a) = −1.
Therefore
−1 1
T 2 (x) = 1 + 0(x − 0) + (x − 0)2 = 1 − x2
2 2
150 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

and
π 1  π 2
T2 =1− ≈ 0.965731
12 2 12
Using a calculator, we find
π
f = 0.965926
12
so
    
T π − f π  = 0.000195
 2 12 12 

33. y = x−2/3 , a = 1, x = 1.2

solution Let f (x) = x−2/3 . Then f  (x) = − 23 x−5/3 , f  (x) = 10 −8/3
9 x , f (1) = 1, f  (1) = − 23 , and f  (1) = 10
9 .
Thus
2 10 2 5
T 2 (x) = 1 − (x − 1) + (x − 1)2 = 1 − (x − 1) + (x − 1)2
3 2·9 3 9
and
2 5 8
T 2 (1.2) = 1 − (0.2) + (0.2)2 = ≈ 0.88889
3 9 9
Using a calculator, f (1.2) = (1.2)−2/3 ≈ 0.88555 so that

|T 2 (1.2) − f (1.2)| ≈ 0.00334

π
34. y = esin x , a= , x = 1.5
2
solution Let f (x) = esin x . Then f  (x) = cos xesin x , f  (x) = cos2 xesin x − sin xesin x , f (a) = e, f  (a) = 0 and
f  (a) = −e. Therefore
 π  −e  π 2 e π 2
T 2 (x) = e + 0 x − + x− =e− x−
2 2 2 2 2
and
e π 2
T 2 (1.5) = e − 1.5 − ≈ 2.711469651
2 2
Using a calculator, we find f (1.5) = 2.711481018, so

|T 2 (1.5) − f (1.5)| = 1.14 × 10−5

√
35. Compute T 3 for f (x) = x centered at a = 1. Then use a plot of the error | f (x) − T 3 (x)| to find a
value c > 1 such that the error on the interval [1, c] is at most 0.25.
solution We have

f (x) = x1/2 f (1) = 1

1 −1/2 1
f  (x) = x f  (1) =
2 2
1 1
f  (x) = − x−3/2 f  (1) = −
4 4
3 3
f  (x) = x−5/2 f  (1) =
8 8
Therefore
1 1 3 1 1 1
T 3 (x) = 1 + (x − 1) − (x − 1)2 + (x − 1)3 = 1 + (x − 1) − (x − 1)2 + (x − 1)3
2 4 · 2! 8 · 3! 2 8 16
A plot of | f (x) − T 3 (x)| is below.
 S E C T I O N 10.7 Taylor Polynomials (LT SECTION 11.7) 151

y
0.25

0.20

0.15

0.10

0.05

x
1.0 1.5 2.0 2.5 3.0

It appears that for x ∈ [1, 2.9] that the error does not exceed 0.25. The error at x = 3 appears to just exceed
0.25.
36. Plot f (x) = 1/(1 + x) together with the Taylor polynomials T n at a = 1 for 1 ≤ n ≤ 4 on the
interval [−2, 8] (be sure to limit the upper plot range).
(a) Over which interval does T 4 appear to approximate f closely?
(b) What happens for x < −1?
(c) Use your computer algebra system to produce and plot T 30 together with f on [−2, 8]. Over which
interval does T 30 appear to give a close approximation?
solution Let f (x) = 1
1+x . Then

1 1
f (x) = f (1) =
1+x 2
1 1
f  (x) = − f  (1) = −
(1 + x)2 4
2 1
f  (x) = f  (1) =
(1 + x)3 4
6 3
f  (x) = − f  (1) = −
(1 + x)4 8
24 3
f (4) (x) = f (4) (1) =
(1 + x)5 4

and
1 1
T 1 (x) = − (x − 1);
2 4
1 1 1
T 2 (x) = − (x − 1) + (x − 1)2 ;
2 4 8
1 1 1 1
T 3 (x) = − (x − 1) + (x − 1)2 − (x − 1)3 ; and
2 4 8 16
1 1 1 1 1
T 4 (x) = − (x − 1) + (x − 1)2 − (x − 1)3 + (x − 1)4
2 4 8 16 32

A plot of f (x), T 1 (x), T 2 (x), T 3 (x) and T 4 (x) is shown below.

y
2 T4 T2

1.5

T1 1

0.5

x
−2 2 T 4 6 8
3

(a) The graph below displays f (x) and T 4 (x) over the interval [−0.5, 2.5]. It appears that T 4 (x) gives a close
approximation to f (x) over the interval (0.1, 2).
152 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

y
1.2
1.0
0.8
0.6
0.4
0.2
x
0.5 1.0 1.5 2.0

(b) For x < −1, f (x) is negative, but the Taylor polynomials are positive; thus, the Taylor polynomials are
poor approximations for x < −1.
(c) The graph below displays f (x) and T 30 (x) over the interval [−2, 8]. It appears that T 30 (x) gives a close
approximation to f (x) over the interval (−1, 3).
y
10

x
−2 2 4 6 8

37. Let T 3 be the Maclaurin polynomial of f (x) = e x . Use the Error Bound to find the maximum possible
value of | f (1.1) − T 3 (1.1)|. Show that we can take K = e1.1 .
solution Since f (x) = e x , we have f (n) (x) = e x for all n; since e x is increasing, the maximum value of e x
on the interval [0, 1.1] is K = e1.1 . Then by the Error Bound,
 1.1  (1.1 − 0)4 e1.1 1.14
e − T 3 (1.1) ≤ K = ≈ 0.183
4! 24
√
38. Let T 2 be the Taylor polynomial of f (x) = x at a = 4. Apply the Error Bound to find the maximum
possible value of the error | f (3.9) − T 2 (3.9)|.
solution We have f (x) = x1/2 , f  (x) = 12 x−1/2 , f  (x) = − 14 x−3/2 , and f  (x) = 38 x−5/2 . This is a decreasing
function of x, so its maximum value on [3.9, 4] is achieved at x = 3.9; that value is 8·3.93 5/2 ≈ 0.0125, so we
can take K = 0.0125. Then
|3.9 − 4|3 0.001
| f (x) − T 2 (x)| ≤ K = 0.0125 ≈ 2.08 × 10−6
3! 6

In Exercises 39–42, compute the Taylor polynomial indicated and use the Error Bound to find the maximum
possible size of the error. Verify your result with a calculator.
39. f (x) = cos x, a = 0; |cos 0.25 − T 5 (0.25)|
solution The Maclaurin series for cos x is

x2 x4 x6
1− + − + ...
2! 4! 6!
so that
x2 x4
T 5 (x) = 1 − +
2 24
T 5 (0.25) ≈ 0.9689127604

In addition, f (6) (x) = − cos x so that | f (6) (x)| ≤ 1 and we may take K = 1 in the Error Bound formula. Then
0.256 1
|cos 0.25 − T 5 (0.25)| ≤ K = 12 ≈ 3.390842014 × 10−7
6! 2 · 6!
(The true value is cos 0.25 ≈ 0.9689124217 and the difference is in fact ≈ 3.387 × 10−7 .)
 S E C T I O N 10.7 Taylor Polynomials (LT SECTION 11.7) 153

40. f (x) = x11/2 , a = 1; | f (1.2) − T 4 (1.2)|

solution Let f (x) = x 11/2
. Then

f (x) = x11/2 f (1) = 1

11 9/2 11
f  (x) = x f  (1) =
2 2
99 7/2 99
f  (x) = x f  (1) =
4 4
693 5/2 693
f  (x) = x f  (1) =
8 8
3465 3/2 3465
f (4) (x) = x f (4) (1) =
16 16
and
11 99 231 1155
T 4 (x) = 1 + (x − 1) + (x − 1)2 + (x − 1)3 + (x − 1)4
2 8 16 128
Using the Error Bound,
K|1.2 − 1|5 K
| f (1.2) − T 4 (1.2)| ≤ =
5! 375000
where K is a number such that | f (5) (x)| ≤ K for x between 1 and 1.2. Now,
10395 1/2
f (5) (x) = x
32
which is increasing for x√> 1. Consequently, on the interval [1, 1.2], f (5) (x) is maximized at x = 1.2. We can
therefore take K = 10395
32 1.2, and then
10395 √
| f (1.2) − T 4 (1.2)| ≤ 1.2 ≈ 9.489 × 10−4
(32)(375000)

41. f (x) = x−1/2 , a = 4; | f (4.3) − T 3 (4.3)|

solution We have
1
f (x) = x−1/2 f (4) =
2
1 1
f  (x) = − x−3/2 f  (4) = −
2 16
3 3
f  (x) = x−5/2 f  (4) =
4 128
15 15
f  (x) = − x−7/2 f  (4) = −
8 1024
105 −9/2
f (4) (x) = x
16
so that
1 1 3 5
T 3 (x) = − (x − 4) + (x − 4)2 − (x − 4)3
2 16 256 2048
Using the Error Bound formula,
|4.3 − 4|4 27K
| f (4.3) − T 3 (4.3)| ≤ K =
4! 80,000
where K is a number such that | f (4) (x)| ≤ K for x between 4 and 4.3. Now, f (4) (x) is a decreasing function
for x > 1, so it takes its maximum value on [4, 4.3] at x = 4; there, its value is
105 −9/2 105
K= 4 =
16 8192
154 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

so that
105
27 8192 27 · 105
| f (4.3) − T 3 (4.3)| ≤ = ≈ 4.3258667 × 10−6
80,000 8192 · 80,000
√ √
42. f (x) = 1 + x, a = 8; | 9.02 − T 3 (8.02)|
√
solution Let f (x) = 1 + x. Then

√
f (x) = 1+x f (8) = 3
1 1
f  (x) = (x + 1)−1/2 f  (8) =
2 6
−1 −1
f  (x) = (x + 1)−3/2 f  (8) =
4 108
3 1
f  (x) = (x + 1)−5/2 f  (8) =
8 648
and
1 1 1 1 1 1
T 3 (x) = 3 + (x − 8) − (x − 8)2 + (x − 8)3 = 3 + (x − 8) − (x − 8)2 + (x − 8)3
6 108 · 2! 648 · 3! 6 216 3888
Therefore
1 1 1
T 3 (8.02) = 3 + (0.02) − (0.02)2 + (0.02)3 = 3.003331484
6 216 3888
Using the Error Bound, we have
√ |8.02 − 8|4 K
| 9.02 − T 3 (8.02)| ≤ K =
4! 150,000,000
where K is a number such that | f (4) (x)| ≤ K for x between 8 and 8.02. Now
15
f (4) (x) = − (1 + x)−7/2
16
which is a decreasing function for 8 ≤ x ≤ 8.02, so we may take
15 −7/2 15
K= 9 =
16 34992
Thus,
√ 15/34992
| 9.02 − T 3 (8.02)| ≤ ≈ 2.858 × 10−12
150,000,000

−1
 polynomial T3 for f (x) = tan x. Compute T 3 2 and use the Error Bound to
1
43. Calculate the Maclaurin
find a bound for the error  tan−1 12 − T 3 12 . Refer to the graph in Figure 8 to find an acceptable value of K.
Verify your result by computing  tan−1 12 − T 3 12  using a calculator.

y
5
4
3
2
1
x
1 2 3
−1

−24x(x2 − 1)
FIGURE 8 Graph of f (4) (x) = , where f (x) = tan−1 x.
(x2 + 1)4
 S E C T I O N 10.7 Taylor Polynomials (LT SECTION 11.7) 155

solution Let f (x) = tan−1 x. Then

f (x) = tan−1 x f (0) = 0

1
f  (x) = f  (0) = 1
1 + x2
−2x
f  (x) = f  (0) = 0
(1 + x2 )2
(1 + x2 )2 (−2) − (−2x)(2)(1 + x2 )(2x)
f  (x) = f  (0) = −2
(1 + x2 )4
and
0 −2 x3
T 3 (x) = 0 + 1(x − 0) + (x − 0)2 + (x − 0)3 = x −
2 6 3
Since f (4) (x) ≤ 5 for x ≥ 0, we may take K = 5 in the Error Bound; then,
    
 −1 1 1  5(1/2)4 5
 tan − T ≤ =
2 
3
2 4! 384

44. Let f (x) = ln(x3 − x + 1). The third Taylor polynomial at a = 1 is

7
T 3 (x) = 2(x − 1) + (x − 1)2 − (x − 1)3
3
Find the maximum possible value of | f (1.1) − T 3 (1.1)|, using the graph in Figure 9 to find an acceptable
value of K. Verify your result by computing | f (1.1) − T 3 (1.1)| using a calculator.
y

41
40

20

x
0.9 1.0 1.1 1.2
FIGURE 9 Graph of f (4) , where f (x) = ln(x3 − x + 1).

solution The maximum value of f (4) (x) on [1.0, 1.1] is less than 41, so we may take K = 41. Then

|1.1 − 1|4 41
| f (1.1) − T 3 (1.1)| ≤ K = ≈ 0.00017083
4! 24 · 10,000
In fact, we have
f (1.1) = ln(1.13 − 1.1 + 1) = ln(1.231) ≈ 0.2078268472
7
T 3 (1.1) = 2(1.1 − 1) + (1.1 − 1)2 − (1.1 − 1)3 ≈ 0.2076666667
3
| f (1.1) − T 3 (1.1)| ≈ 0.2078268472 − 0.2076666667 = 0.0001601805
which is in accordance with the Error Bound above.
45. Let T 2 be the Taylor polynomial at a = 0.5 for f (x) = cos(x2 ). Use the error bound to find the
maximum possible value of | f (0.6) − T 2 (0.6)|. Plot f (3) to find an acceptable value of K.
solution We have

f (x) = cos(x2 ) f (0.5) = cos(0.25) ≈ 0.9689


f (x) = −2x sin(x ) 2
f  (0.5) = − sin(0.25) ≈ −0.2474039593
f  (x) = −4x2 cos(x2 ) − 2 sin(x2 ) f  (0.5) = − cos(0.25) − 2 sin(0.25) ≈ −1.463720340
f (3) (x) = 8x3 sin(x2 ) − 12x cos(x2 )
156 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

so that

T 2 (x) = 0.9689 − 0.2472039593(x − .5) − 0.73186017(x − .5)2

and T 2 (0.6) ≈ 0.9368534237. A graph of f (3) (x) for x near 0.5 is below.
0.4 0.5 0.6 0.7
−3 x

−4

−5

−6

−7
y

Clearly the maximum value of | f (3) (x)| on [0.5, 0.6] is bounded by 7 (near x = 0.5), so we may take K = 7;
then
|0.6 − 0.5|3 7
| f (0.6) − T 2 (0.6)| ≤ K = ≈ 0.0011666667
3! 6000

46. Calculate the Maclaurin

 polynomial
 T 2 for f (x) = sech x and use the Error Bound to find the
maximum possible value of  f 12 − T 2 12 . Plot f  to find an acceptable value of K.
solution To compute T 2 (x) for f (x) = sech x, we take the first two derivatives:

f (x) = sech x f (0) = 1


f (x) = − sech x tanh x f  (0) = 0
f  (x) = sech x tanh2 x − sech3 x f  (0) = −1

From this,
1 2
T 2 (x) = 1 − x ,
2
and
   2
1 1 1 1 7
T2 =1− =1− = .
2 2 2 8 8
Using the Error Bound, we have
    
 1 1  |1/2|3 K
 f − T2  ≤ K = ,
2 2 6 48

where K is a number such that | f  (x)| ≤ K| for x between 0 and 12 . Here,

f  (x) = − sech x tanh3 x + 2 sech3 x tanh x + 3 sech2 x(sech x tanh x)

= 5 sech2 x tanh x − sech x tanh3 x.

A plot of f  (x) is given below. From the plot, we see that | f  (x)| ≤ 2 for all x between 0 and 1/2. Thus,
    
 1 1  2 1
 f − T2 ≤ = .
2 2  48 24
y
2
1.5
1
0.5
x
0 0.1 0.2 0.3 0.4 0.5
 S E C T I O N 10.7 Taylor Polynomials (LT SECTION 11.7) 157

In Exercises 47–50, use the Error Bound to find a value of n for which the given inequality is satisfied. Then
verify your result using a calculator.

47. | cos 0.1 − T n (0.1)| ≤ 10−7 , a=0

solution Using the Error Bound with K = 1 (every derivative of f (x) = cos x is ± sin x or ± cos x, so
| f (n) (x)| ≤ 1 for all n), we have

(0.1)n+1
|T n (0.1) − cos 0.1| ≤
(n + 1)!

With n = 3,

(0.1)4
≈ 4.17 × 10−6 > 10−7
4!
but with n = 4,

(0.1)5
≈ 8.33 × 10−8 < 10−7
5!
so we choose n = 4. Now,
1 2 1 4
T 4 (x) = 1 − x + x
2 24
so
1 1
T 4 (0.1) = 1 − (0.1)2 + (0.1)4 = 0.995004166
2 24
Using a calculator, cos 0.1 = 0.995004165, so

|T 4 (0.1) − cos 0.1| = 1.387 × 10−8 < 10−7

48. | ln 1.3 − T n (1.3)| ≤ 10−4 , a=1

solution Let f (x) = ln x. Then f  (x) = x−1 , f  (x) = −x−2 , f  (x) = 2x−3 , f (4) (x) = −6x−4 , etc. In general,

f (n) (x) = (−1)n+1 (n − 1)!x−n

Now, | f (n+1) (x)| is decreasing on the interval [1, 1.3], so | f (n+1) (x)| ≤ | f (n+1) (1)| = n! for all x ∈ [1, 1.3]. We
can therefore take K = n! in the Error Bound, and

|1.3 − 1|n+1 (0.3)n+1

| ln 1.3 − T n (1.3)| ≤ n! =
(n + 1)! n+1

With n = 5,

(0.3)6
= 1.215 × 10−4 > 10−4
6
but with n = 6,

(0.3)7
= 3.124 × 10−5 < 10−4
7
Therefore, the error is guaranteed to be below 10−4 for n = 6. Now,

1 1 1 1 1
T 6 (x) = (x − 1) − (x − 1)2 + (x − 1)3 − (x − 1)4 + (x − 1)5 − (x − 1)6
2 3 4 5 6
and T 6 (1.3) ≈ 0.2623395000. Using a calculator, ln(1.3) ≈ 0.2623642645; the difference is

ln(1.3) − T 6 (1.3) ≈ 0.0000247645 < 10−4

158 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

√
49. | 1.3 − T n (1.3)| ≤ 10−6 , a=1

solution Using the Error Bound, we have

√ |1.3 − 1|n+1 |0.3|n+1
| 1.3 − T n (1.3)| ≤ K =K
(n + 1)! (n + 1)!
√
where K is a number such that | f (n+1) (x)| ≤ K for x between 1 and 1.3. For f (x) = x, | f (n) (x)| is decreasing
for x > 1, hence the maximum value of | f (n+1) (x)| occurs at x = 1. We may therefore take
1 · 3 · 5 · · · (2n + 1)
K = | f (n+1) (1)| =
2n+1
1 · 3 · 5 · · · (2n + 1) 2 · 4 · 6 · · · (2n + 2) (2n + 2)!
= · =
2n+1 2 · 4 · 6 · · · (2n + 2) (n + 1)!22n+2
Then
√ (2n + 2)! |.3|n+1 (2n + 2)!
| 1.3 − T n (1.3)| ≤ · = (0.075)n+1
(n + 1)!22n+2 (n + 1)! [(n + 1)!]2
With n = 9
(20)!
(0.075)10 = 1.040 × 10−6 > 10−6
[(10)!]2
but with n = 10
(22)!
(0.075)11 = 2.979 × 10−7 < 10−6
[(11)!]2
Hence, n = 10 will guarantee the desired accuracy. Using technology to compute and evaluate T 10 (1.3) gives
√
T 10 (1.3) ≈ 1.140175414, 1.3 ≈ 1.140175425
and
√
| 1.3 − T 10 (1.3)| ≈ 1.1 × 10−8 < 10−6

50. |e−0.1 − T n (−0.1)| ≤ 10−6 , a=0

solution Using the Error Bound, we have

|−0.1 − 0|n+1 1
|e−0.1 − T n (−0.1)| ≤ K = K n+1
(n + 1)! 10 (n + 1)!
where K is a number such that | f (n+1) (x)| ≤ K for x between −0.1 and 0. Since f (x) = e x , f (n) (x) = e x for
all n; this is an increasing function, so it takes its maximum value at x = 0; this value is 1. So we may take
K = 1 and then
1
|e−0.1 − T n (−0.1)| ≤
10n+1 (n + 1)!
With n = 3
1 1
= ≈ 4.166666667 × 10−6 > 10−6
104 · 24 240,000
but with n = 4
1 1
= ≈ 8.333333333 × 10−8 < 10−6
105 · 120 12,000,000
Thus n = 4 will guarantee the desired accuracy. Using technology to compute T 4 (x) and evalute,
T 4 (−0.1) ≈ 0.9048375000, e−0.1 ≈ 0.9048374180
and
|e−0.1 − T 4 (−0.1)| ≈ 8.2 × 10−8 < 10−6
 S E C T I O N 10.7 Taylor Polynomials (LT SECTION 11.7) 159

x2 x3
51. Let f (x) = e−x and T 3 (x) = 1 − x + − .
2 6
(a) Use the Error Bound to show that for all x ≥ 0,
x4
| f (x) − T 3 (x)| ≤
24
(b) Illustrate this inequality by plotting y = f (x) − T 3 (x) and y = x4 /24 together over [0, 1].

solution
(a) Note that f (n) (x) = ±e−x , so that | f (n) (x)| = f (x). Now, f (x) is a decreasing function for x ≥ 0, so that for
any c > 0, | f (n) (x)| takes its maximum value at x = 0; this value is e0 = 1. Thus we may take K = 1 in the
Error Bound equation. Thus for any x,
|x − 0|4 x4
| f (x) − T 3 (x)| ≤ K =
4! 24
x4
(b) A plot of f (x) − T 3 (x) and 24 is shown below.
y
10

1 4
8 x
24

x
1 2 3 4 5 6 7

52. Use the Error Bound with n = 4 to show that

  
 x3  |x|5
sin x − x − ≤ (for all x)
6  120
solution Note that all derivatives of sin x are either ± cos x or ± sin x so are bounded in absolute value by
1. Thus we may take K = 1 in the Error Bound. Now,
x3
T 4 (x) = x −
3!
so that
  
 x3  |x − 0|5 |x|5
|sin x − T 4 (x)| = sin x − x −  ≤K =
 6  5! 120

53. Let T n be the Taylor polynomial for f (x) = ln x at a = 1, and let c > 1. Show that
|c − 1|n+1
| ln c − T n (c)| ≤
n+1
Then find a value of n such that | ln 1.5 − T n (1.5)| ≤ 10−2 .

solution With f (x) = ln x, we have

f  (x) = x−1 , f  (x) = −x−2 , f  (x) = 2x−3 , f (4) (x) = −6x−4
and, in general,
f (k+1) (x) = (−1)k k! x−k−1
Notice that | f (k+1) (x)| = k!|x|−k−1 is a decreasing function for x > 0. Therefore, the maximum value of
| f (k+1) (x)| on [1, c] is | f (k+1) (1)|. Using the Error Bound, we have
|c − 1|n+1
|ln c − T n (c)| ≤ K
(n + 1)!
160 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

where K is a number such that | f (n+1) (x)| ≤ K for x between 1 and c. We have just determined that we may
take K = | f (n+1) (1)| = n!. Then
|c − 1|n+1 |c − 1|n+1
|ln c − T n (c)| ≤ n! =
(n + 1)! n+1
Evaluating at c = 1.5 gives
|1.5 − 1|n+1 (0.5)n+1
|ln 1.5 − T n (1.5)| ≤ =
n+1 n+1
With n = 3,
(0.5)4
= 0.015625 > 10−2
4
but with n = 4,
(0.5)5
= 0.00625 < 10−2
5
Hence, n = 4 will guarantee the desired accuracy.

54. Let n ≥ 1. Show that if |x| is small, then

x 1−n 2
(x + 1)1/n ≈ 1 + + x
n 2n2
Use this approximation with n = 6 to estimate 1.51/6 .

solution Let f (x) = (x + 1)1/n . Then

f (x) = (x + 1)1/n f (0) = 1
1 1
f  (x) = (x + 1)1/n−1 f  (0) =
n n
   
1 1 1 1
f  (x) = − 1 (x + 1)1/n−2 f  (0) = −1
n n n n
and
   
1 1 1 x2 x 1−n 2
T 2 (x) = 1 + (x) + 2 − =1+ + x
n n n 2 n 2n2
With n = 6 and x = 0.5,
307
1.51/6 ≈ T 2 (0.5) = ≈ 1.065972
288
55. Verify that the third Maclaurin polynomial for f (x) = e x sin x is equal to the product of the third Maclau-
rin polynomials of f (x) = e x and f (x) = sin x (after discarding terms of degree greater than 3 in the product).

solution Let f (x) = e x sin x. Then

f (x) = e x sin x f (0) = 0

f (x) = e (cos x + sin x)
x
f  (0) = 1
f  (x) = 2e x cos x f  (0) = 2
f  (x) = 2e x (cos x − sin x) f  (0) = 2
and
2 2 2 3 x3
T 3 (x) = 0 + (1)x + x + x = x + x2 +
2! 3! 3
2 3
Now, the third Maclaurin polynomial for e x is 1 + x + x2 + x6 , and the third Maclaurin polynomial for sin x
3
is x − x6 . Multiplying these two polynomials, and then discarding terms of degree greater than 3, yields
x3
e x sin x ≈ x + x2 +
3
which agrees with the Maclaurin polynomial obtained from the definition.
 S E C T I O N 10.7 Taylor Polynomials (LT SECTION 11.7) 161

56. Find the fourth Maclaurin polynomial for f (x) = sin x cos x by multiplying the fourth Maclaurin poly-
nomials for f (x) = sin x and f (x) = cos x.
3
solution The fourth Maclaurin polynomial for sin x is x − x6 , and the fourth Maclaurin polynomial for
2
x4
cos x is 1 − x2 + 24 . Multiplying these two polynomials, and then discarding terms of degree greater than 4,
we find that the fourth Maclaurin polynomial for f (x) = sin x cos x is
2x3
T 4 (x) = x −
3

57. Find the Maclaurin polynomials T n for f (x) = cos(x2 ). You may use the fact that T n (x) is equal to the
sum of the terms up to degree n obtained by substituting x2 for x in the nth Maclaurin polynomial of cos x.

solution The Maclaurin polynomials for cos x are of the form

x2 x4 x2n
T 2n (x) = 1 − + + · · · + (−1)n
2 4! (2n)!
Accordingly, the Maclaurin polynomials for cos(x2 ) are of the form
x4 x8 x4n
T 4n (x) = 1 − + + · · · + (−1)n
2 4! (2n)!

58. Find the Maclaurin polynomials of 1/(1 + x2 ) by substituting −x2 for x in the Maclaurin polynomials of
1/(1 − x).

solution The Maclaurin polynomials for 1

1−x are of the form

T n (x) = 1 + x + x2 + · · · + xn
1
Accordingly, the Maclaurin polynomials for 1+x2
are of the form

T 2n (x) = 1 − x2 + x4 − x6 + · · · + (−x2 )n

59. Let f (x) = 3x3 + 2x2 − x − 4. Calculate T j for j = 1, 2, 3, 4, 5 at both a = 0 and a = 1. Show that
T 3 (x) = f (x) in both cases.

solution Let f (x) = 3x3 + 2x2 − x − 4. Then

f (x) = 3x3 + 2x2 − x − 4 f (0) = −4 f (1) = 0

 
f (x) = 9x + 4x − 1
2
f (0) = −1 f  (1) = 12
f  (x) = 18x + 4 f  (0) = 4 f  (1) = 22
f  (x) = 18 f  (0) = 18 f  (1) = 18
f (4) (x) = 0 f (4) (0) = 0 f (4) (1) = 0
f (5) (x) = 0 f (5) (0) = 0 f (5) (1) = 0
At a = 0,
T 1 (x) = −4 − x;
T 2 (x) = −4 − x + 2x2 ;
T 3 (x) = −4 − x + 2x2 + 3x3 = f (x);
T 4 (x) = T 3 (x); and
T 5 (x) = T 3 (x)
At a = 1,
T 1 (x) = 12(x − 1);
T 2 (x) = 12(x − 1) + 11(x − 1)2 ;
162 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

T 3 (x) = 12(x − 1) + 11(x − 1)2 + 3(x − 1)3 = −4 − x + 2x2 + 3x3 = f (x);

T 4 (x) = T 3 (x); and
T 5 (x) = T 3 (x)

60. Let T n be the nth Taylor polynomial at x = a for a polynomial f of degree n. Based on the result of
Exercise 59, guess the value of | f (x) − T n (x)|. Prove that your guess is correct using the Error Bound.

solution Based on Exercise 59, we expect | f (x) − T n (x)| = 0. From the Error Bound,

|x − a|n+1
| f (x) − T n (x)| ≤ K
(n + 1)!
where K is a number such that | f (n+1) (u)| ≤ K for u between a and x. Since f (n+1) (x) = 0 for an nth degree
polynomial, we may take K = 0; the Error Bound then becomes | f (x) − T n (x)| = 0.

61. Let s(t) be the distance of a truck to an intersection. At time t = 0, the truck is 60 m from the intersection,
travels away from it with a velocity of 24 m/s, and begins to slow down with an acceleration of a = −3
m/s2 . Determine the second Maclaurin polynomial of s, and use it to estimate the truck’s distance from the
intersection after 4 s.

solution Place the origin at the intersection, so that s(0) = 60 (the truck is traveling away from the inter-
section). The second Maclaurin polynomial of s(t) is
s (0) 2
T 2 (t) = s(0) + s (0)t + t
2
The conditions of the problem tell us that s(0) = 60, s (0) = 24, and s (0) = −3. Thus
3
T 2 (t) = 60 + 24t − t2
2
so that after 4 seconds,
3 2
T 2 (4) = 60 + 24 · 4 − · 4 = 132 m
2
The truck is 132 m past the intersection.

62. A bank owns a portfolio of bonds whose value P(r) depends on the interest rate r (measured in percent;
e.g., r = 5 means a 5% interest rate). The bank’s quantitative analyst determines that
 
dP  d2 P 
P(5) = 100,000,  = −40,000,  = 50,000
dr r=5 dr2 r=5
In finance, this second derivative is called bond convexity. Find the second Taylor polynomial of P(r) cen-
tered at r = 5 and use it to estimate the value of the portfolio if the interest rate moves to r = 5.5%.

solution The second Taylor polynomial of P(r) at r = 5 is

P (5)
T 2 (r) = P(5) + P (5)(r − 5) + (r − 5)2
2
From the conditions of the problem, P(5) = 100,000, P (5) = −40,000, and P (5) = 50,000, so that
T 2 (r) = 100,000 − 40,000(r − 5) + 25,000(r − 5)2
If the interest rate moves to 5.5%, then the value of the portfolio can be estimated by
T 2 (5.5) = 100,000 − 40,000(0.5) + 25,000(0.5)2 = 86,250

63. A narrow, negatively charged ring of radius R exerts a force on a positively charged particle P located at
distance x above the center of the ring of magnitude
kx
F(x) = −
(x2 + R2 )3/2
where k > 0 is a constant (Figure 10).
 S E C T I O N 10.7 Taylor Polynomials (LT SECTION 11.7) 163

(a) Compute the third-degree Maclaurin polynomial for F.

(b) Show that F ≈ −(k/R3 )x to second order. This shows that when x is small, F(x) behaves like a restoring
force similar to the force exerted by a spring.
(c) Show that F(x) ≈ −k/x2 when x is large by showing that
F(x)
lim =1
x→∞ −k/x2
Thus, F(x) behaves like an inverse square law, and the charged ring looks like a point charge from far away.

F(x)
R
x
Nearly linear
here
x
P Nearly inverse square
here

FIGURE 10

solution
(a) Start by computing and evaluating the necessary derivatives:
kx
F(x) = − F(0) = 0
(x2 + R2 )3/2
k(2x2 − R2 ) k
F  (x) = F  (0) = −
(x2 + R2 )5/2 R3
3kx(3R2 − 2x2 )
F  (x) = F  (0) = 0
(x2 + R2 )7/2
3k(8x4 − 24x2 R2 + 3R4 ) 9k
F  (x) = F  (0) =
(x2 + R2 )9/2 R5
so that
F  (0) 2 F  (0) 3 k 3k
T 3 (x) = F(0) + F  (0)x + x + x = − 3 x + 5 x3
2! 3! R 2R
(b) To degree 2, F(x) ≈ T 3 (x) ≈ − Rk3 x as we may ignore the x3 term of T 3 (x).
(c) We have
 2 
F(x) x −kx x3
lim = lim − · = lim
x→∞ −k/x2 x→∞ k (x2 + R2 )3/2 x→∞ (x2 + R2 )3/2

1 1
= lim = lim
x→∞ x−3 (x2 + R2 )3/2 x→∞ (1 + R2 /x2 )3/2
=1
Thus as x grows large, F(x) looks like an inverse square function.

64. A light wave of wavelength λ travels from A to B by passing through an aperture (circular region) located
in a plane that is perpendicular to AB (see Figure 11 for the notation). Let f (r) = d + h ; that is, f (r) is the
distance AC + CB as a function of r.
√ √
(a) Show that f (r) = d2 + r2 + h2 + r2 , and use the Maclaurin polynomial of order 2 to show that
 
1 1 1 2
f (r) ≈ d + h + + r
2 d h
(b) The Fresnel zones, used to determine the optical disturbance at B, are
√ the concentric bands bounded by
the circles of radius Rn such that f (Rn ) = d + h + nλ/2. Show that Rn ≈ nλL, where L = (d−1 + h−1 )−1 .
(c) Estimate the radii R1 and R100 for blue light (λ = 475 × 10−7 cm) if d = h = 100 cm.
164 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

C
d9
A r
d h9
O

R1 h
R2 B
R3

FIGURE 11 The Fresnel zones are the regions between the circles of radius Rn .

solution
√ √ √
(a) From the diagram, we see that AC =
√ d2 + r2 and CB = h2 + r2 . Therefore, f (r) = d2 + r2 +
h2 + r2 . Moreover,
r r d2 h2
f  (r) = √ + √ , f  (r) = + 2
d +r
2 2 h + r2
2 (d2 +r )
2 3/2 (h + r2 )3/2

f (0) = d + h, f  (0) = 0 and f  (0) = d−1 + h−1 . Thus,

 
1 1 1 2
f (r) ≈ T 2 (r) = d + h + + r
2 d h
(b) Solving
 
1 1 1 2 nλ
f (Rn ) ≈ d + h + + R =d+h+
2 d h n 2
yields
 √
Rn = nλ(d−1 + h−1 )−1 = nλL
where L = (d−1 + h−1 )−1 .
(c) With d = h = 100 cm, L = 50 cm. Taking λ = 475 × 10−7 cm, it follows that
√
R1 ≈ λL = 0.04873 cm; and
√
R100 ≈ 100λL = 0.4873 cm

65. Referring to Figure 12, let a be the length of the chord AC of angle θ of the unit circle. Derive the
following approximation for the excess of the arc over the chord:
θ3
θ−a≈
24
Hint: Show that θ − a = θ − 2 sin(θ/2) and use the third Maclaurin polynomial as an approximation.

a B
b

2
A
FIGURE 12 Unit circle.

solution Draw a line from the center O of the circle to B, and label the point of intersection of this line
with AC as D. Then CD = a2 , and the angle COB is 2θ . Since CO = 1, we have
θ a
sin =
2 2
 S E C T I O N 10.7 Taylor Polynomials (LT SECTION 11.7) 165

so that a = 2 sin(θ/2). Thus θ − a = θ − 2 sin(θ/2). Now, the third Maclaurin polynomial for f (θ) = sin(θ/2)
can be computed as follows: f (0) = 0, f  (x) = 12 cos(θ/2) so that f  (0) = 12 . f  (x) = − 14 sin(θ/2) and
f  (0) = 0. Finally, f  (x) = − 18 cos(θ/2) and f  (0) = − 18 . Thus

f  (0) 2 f  (0) 3 1 1

T 3 (θ) = f (0) + f  (0)θ + θ + θ = θ − θ3
2! 3! 2 48
Finally,
 
θ 1 3 θ3
θ − a = θ − 2 sin ≈ θ − 2T 3 (θ) = θ − θ − θ =
2 24 24

66. To estimate the length θ of a circular arc of the unit circle, the seventeenth-century Dutch scientist
Christian Huygens used the approximation θ ≈ (8b − a)/3, where a is the length of the chord AC of angle θ
and b is the length of the chord AB of angle θ/2 (Figure 12).
(a) Prove that a = 2 sin(θ/2) and b = 2 sin(θ/4), and show that the Huygens approximation amounts to the
approximation
16 θ 2 θ
θ≈ sin − sin
3 4 3 2
(b) Compute the fifth Maclaurin polynomial of the function on the right.
(c) Use the Error Bound to show that the error in the Huygens approximation is less than 0.00022|θ|5 .

solution
(a) By the Law of Cosines and the identity sin2 (θ/2) = (1 − cos θ)/2:
θ
a2 = 12 + 12 − 2 cos θ = 2(1 − cos θ) = 4 sin2
2
and so a = 2 sin(θ/2). Similarly, b = 2 sin(θ/4). Substituting these expressions for a and b into the Huygens
approximation yields
8 θ 1 θ 16 θ 2 θ
θ≈ · 2 sin − · 2 sin = sin − sin
3 4 3 2 3 4 3 2
x3 x5
(b) The fifth Maclaurin polynomial for sin x is x − 6 + 120 ; therefore, the fifth Maclaurin polynomial for
sin(θ/2) is

θ (θ/2)3 (θ/2)5 θ θ3 θ5
− + = − +
2 6 120 2 48 3840
and the fifth Maclaurin polynomial for sin(θ/4) is

θ (θ/4)3 (θ/4)5 θ θ3 θ5
− + = − +
4 6 120 4 384 122880
Thus, the fifth Maclaurin polynomial for f (θ) = 16
3 sin 4θ − 23 sin 2θ is

1 5
θ− θ
7680
(c) Based on the result from part (b), the Huygens approximation for θ is equal to the fourth Maclaurin
polynomial T 4 (θ) for f (θ), and the error is at most K|θ|5 /5!, where K is the maximum value of the absolute
value of the fifth derivative f (5) (θ). Because
1 θ 1 θ
f (5) (θ) = cos − cos
192 4 48 2
we may take K = 1/48 + 1/192 = 0.0260417, so the error is at most |θ|5 times the constant
0.0261
= 0.00022
5!
166 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

Further Insights and Challenges

67. Show that the nth Maclaurin polynomial of f (x) = arcsin x for n odd is

1 x3 1 · 3 x5 1 · 3 · 5 · · · (n − 2) xn
T n (x) = x + + + ··· +
2 3 2·4 5 2 · 4 · 6 · · · (n − 1) n

solution Let f (x) = sin−1 x. Then

f (x) = sin−1 x f (0) = 0

1
f  (x) = √ f  (0) = 1
1 − x2
1
f  (x) = − (1 − x2 )−3/2 (−2x) f  (0) = 0
2
2x 2
+1
f  (x) = f  (0) = 1
(1 − x2 )5/2
−3x(2x2 + 3)
f (4) (x) = f (4) (0) = 0
(1 − x2 )7/2
24x4 + 72x2 + 9
f (5) (x) = f (5) (0) = 9
(1 − x2 )9/2
.. ..
. .
f (7) (0) = 225

and
x3 9x5 225x7 1 x3 1 3 x5 1 3 5 x7
T 7 (x) = x + + + = x+ + +
3! 5! 7! 2 3 24 5 246 7
Thus, we can infer that
1 x3 1 3 x5 1 3 5 x7 1 3 n − 2 xn
T n (x) = x + · + + + ··· + ···
2 3 2 4 5 2 46 7 2 4 n−1 n

68. Let x ≥ 0 and assume that f (n+1) (t) ≥ 0 for 0 ≤ t ≤ x. Use Taylor’s Theorem to show that the nth
Maclaurin polynomial T n satisfies

T n (x) ≤ f (x), for all x ≥ 0

solution From Taylor’s Theorem,

x
1
Rn (x) = f (x) − T n (x) = (x − u)n f (n+1) (u) du
n! 0

If f (n+1) (t) ≥ 0 for all t then

x
1
(x − u)n f (n+1) (u) du ≥ 0
n! 0

since (x − u) ≥ 0 for 0 ≤ u ≤ x. Thus, f (x) − T n (x) ≥ 0, or f (x) ≥ T n (x).

n

69. Use Exercise 68 to show that for x ≥ 0 and all n,

x2 xn
ex ≥ 1 + x + +···+
2! n!
Sketch the graphs of y = e x , y = T 1 (x), and y = T 2 (x) on the same coordinate axes. Does this inequality
remain true for x < 0?
solution Let f (x) = e x . Then f (n) (x) = e x for all n. Because e x > 0 for all x, it follows from Exercise 68
that f (x) ≥ T n (x) for all x ≥ 0 and for all n. For f (x) = e x ,

x2 xn
T n (x) = 1 + x + +···+
2! n!
 S E C T I O N 10.7 Taylor Polynomials (LT SECTION 11.7) 167

thus,
x2 xn
ex ≥ 1 + x + +···+
2! n!
From the figure below, we see that the inequality does not remain true for x < 0, as T 2 (x) ≥ e x for x < 0.
y ex

6
T2
4
T1
2

x
−2 −1 1 2

70. This exercise is intended to reinforce the proof of Taylor’s Theorem.

x
(a) Show that f (x) = T 0 (x) + f  (u) du.
a
(b) Use Integration by Parts to prove the formula
x x
(x − u) f (2) (u) du = − f  (a)(x − a) + f  (u) du
a a

(c) Prove the case n = 2 of Taylor’s Theorem:

x
f (x) = T 1 (x) + (x − u) f (2) (u) du
a

solution
(a)
x
T 0 (x) + f  (u) du = T 0 (x) + f (x) − f (a) (from FTC2)
a
= f (a) + f (x) − f (a) = f (x)
(b) Using Integration by Parts with w = x − u and v = f  (u) du,
x  x x
(x − u) f  (u) du = f  (u)(x − u) + f  (u) du
a a a
x
= f  (x)(x − x) − f  (a)(x − a) + f  (u) du
a
x
= − f  (a)(x − a) + f  (u) du
a

(c)
x x
T 1 (x) + (x − u) f  (u) du = f (a) + f  (a)(x − a) + − f  (a)(x − a) + f  (u) du
a a
= f (a) + f (x) − f (a) = f (x)

In Exercises 71–75, we estimate integrals using Taylor polynomials. Exercise 72 is used to estimate the error.
1/2
71. Find the fourth Maclaurin polynomial T 4 for f (x) = e−x , and calculate I =
2
T 4 (x) dx as an estimate
0
1/2
e−x dx. A CAS yields the value I ≈ 0.461281. How large is the error in your approximation? Hint:
2
for
0
T 4 is obtained by substituting −x2 in the second Maclaurin polynomial for e x .
solution Following the hint, since the second Maclaurin polynomial for e x is

x2
1+x+
2
168 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

2
we substitute −x2 for x to get the fourth Maclaurin polynomial for e x :
x4
T 4 (x) = 1 − x2 +
2
Then
 
1/2
−x2
1/2
1 3 1 5 1/2 443
e dx ≈ T 4 (x) dx = x − x + x  = ≈ 0.4614583333
0 0 3 10 0 960
$ 1/2
e−x dx ≈ 0.4612810064, so the error is about 1.77 × 10−4 .
2
Using a CAS, we have 0

72. Approximating Integrals Let L > 0. Show that if two functions f and g satisfy | f (x) − g(x)| < L for
all x ∈ [a, b], then
 
 b b

 f (x) dx − g(x) dx dx < L(b − a)
a a 

solution Because f (x) − g(x) ≤ | f (x) − g(x)|, it follows that

 b b   b  b
 f (x) dx − g(x) dx =  ( f (x) − g(x)) dx ≤ | f (x) − g(x)| dx
a a a a
b
< L dx = L(b − a)
a

73. Let T 4 be the fourth Maclaurin polynomial for f (x) = cos x.

(a) Show that
 6 % &
1
2 1
| cos x − T 4 (x)| ≤ for all x ∈ 0,
6! 2
Hint: T 4 (x) = T 5 (x).
1/2 1/2
(b) Evaluate T 4 (x) dx as an approximation to cos x dx. Use Exercise 72 to find a bound for the
0 0
size of the error.
solution
(a) Let f (x) = cos x. Then
x2 x4
T 4 (x) = 1 − +
2 24
Moreover, with a = 0, T 4 (x) = T 5 (x) and
|x|6
|cos x − T 4 (x)| ≤ K
6!
where K is a number such that | f (6) (u)| ≤ K for u between 0 and x. Now | f (6) (u)| = | cos u| ≤ 1, so we may
take K = 1. Finally, with the restriction x ∈ [0, 12 ],

(1/2)6
|sin x − T 4 (x)| ≤ ≈ 0.000022
6!
(b)
1/2  
x2 x4 1841
1− + dx = ≈ 0.479427
0 2 24 3840
By (a) and Exercise 72, the error associated with this approximation is less than or equal to
 
(1/2)6 1 1
−0 = ≈ 1.1 × 10−5
6! 2 92160
1/2
Note that cos x dx ≈ 0.4794255, so the actual error is roughly 1.5 × 10−6 .
0
 S E C T I O N 10.7 Taylor Polynomials (LT SECTION 11.7) 169

74. Let Q(x) = 1 − x2 /6. Use the Error Bound for f (x) = sin x to show that
 
 sin x − Q(x) ≤ |x|
4
 x  5!
1 1
Then calculate Q(x) dx as an approximation to (sin x/x) dx and find a bound for the error.
0 0

solution The third Maclaurin polynomial for sin x is

1 3 1
T 3 (x) = x − x = x − x3 = xQ(x)
3! 6
Additionally, this is also T 4 (x) since (sin x)(4) (0) = 0. All derivatives of sin x are either ± sin x or ± cos x,
which are bounded in absolute value by 1. Thus we may take K = 1 in the Error Bound, so
|x|5 |x|5
|sin x − xQ(x)| = |sin x − T 3 (x)| = |sin x − T 4 (x)| ≤ K =
5! 5!
Divide both sides of this inequality by |x| to get
 
 sin x − Q(x) ≤ |x|
4
 x  5!
$1
We can thus estimate 0 (sin x/x) dx by
1   
1
x2 x3 1 17
Q(x) dx = 1− dx = x −  = ≈ 0.9444444444
0 0 6 18 0 18
The error in this approximation is at most
|1|4 1
= ≈ 0.008333333333
5! 120
The true value of the integral is approximately 0.9460830704, which is consistent with the Error Bound.
75. (a) Compute the sixth Maclaurin polynomial T 6 for f (x) = sin(x2 ) by substituting x2 in P(x) = x − x3 /6,
the third Maclaurin polynomial for f (x) = sin x.
|x|10
(b) Show that | sin(x2 ) − T 6 (x)| ≤ .
5!
Hint: Substitute x for x in the Error Bound for | sin x − P(x)|, noting that P is also the fourth Maclaurin
2

polynomial for f (x) = sin x.

1/2
(c) Use T 6 to approximate sin(x2 ) dx and find a bound for the error.
0

solution Let s(x) = sin x and f (x) = sin(x2 ). Then

(a) The third Maclaurin polynomial for sin x is
x3
S 3 (x) = x −
6
so, substituting x2 for x, we see that the sixth Maclaurin polynomial for sin(x2 ) is
x6
T 6 (x) = x2 −
6
(b) Since all derivatives of s(x) are either ± cos x or ± sin x, they are bounded in magnitude by 1, so we may
take K = 1 in the Error Bound for sin x. Since the third Maclaurin polynomial S 3 (x) for sin x is also the
fourth Maclaurin polynomial S 4 (x), we have
|x|5 |x|5
|sin x − S 3 (x)| = |sin x − S 4 (x)| ≤ K =
5! 5!
Now substitute x2 for x in the above inequality and note from part (a) that S 3 (x2 ) = T 6 (x) to get
|x2 |5 |x|10
|sin(x2 ) − S 3 (x2 )| = |sin(x2 ) − T 6 (x)| ≤ =
5! 5!
170 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

(c)
 
1/2 1/2
1 3 1 7 1/2
sin(x2 ) dx ≈ T 6 (x) dx = x − x  ≈ 0.04148065476
0 0 3 42 0

From part (b) and Exercise 72, the error is bounded by

 
x10 1 (1/2)11 1
−0 = = ≈ 4.069010417 × 10−6
5! 2 120 2048 · 120
The true value of the integral is approximately 0.04148102427, which is consistent with the computed Error
Bound.
76. Prove by induction that for all k,
 
d j (x − a)k k(k − 1) · · · (k − j + 1)(x − a)k− j
j
=
dx k! k!
  ⎧
d j (x − a)k  ⎪
⎪
⎨1 for k = j
=
dx j k!  x=a ⎪
 ⎪
⎩0 for k  j

Use this to prove that T n agrees with f at x = a to order n.

solution The first formula is clearly true for j = 0. Suppose the formula is true for an arbitrary j. Then
     
d j+1 (x − a)k d d j (x − a)k d k(k − 1) · · · (k − j + 1)(x − a)k− j
= =
dx j+1 k! dx dx j k! dx k!
k(k − 1) · · · (k − j + 1)(k − ( j + 1) + 1)(x − a)k−( j+1)
=
k!
as desired. Note that if k = j, then the numerator is k!, the denominator is k! and the value of the derivative
is 1; otherwise, the value of the derivative is 0 at x = a. In other words,
  ⎧
d j (x − a)k  ⎪
⎪
⎨1 for k = j

 = ⎪
⎪
dx j k! x=a ⎩0 for k  j

Applying this latter formula, it follows that

   
dj  =
n
d j f (k) (a) 
k 
j
T n (a)  j
(x − a)  x=a = f (a)
( j)
dx x=a k=0
dx k!

as required.
77. Let a be any number and let

P(x) = an xn + an−1 xn−1 + · · · + a1 x + a0

be a polynomial of degree n or less.

(a) Show that if P( j) (a) = 0 for j = 0, 1, . . . , n, then P(x) = 0, that is, a j = 0 for all j. Hint: Use induction,
noting that if the statement is true for degree n − 1, then P (x) = 0.
(b) Prove that T n is the only polynomial of degree n or less that agrees with f at x = a to order n. Hint: If Q
is another such polynomial, apply (a) to P(x) = T n (x) − Q(x).
solution
(a) Note first that if n = 0, i.e., if P(x) = a0 is a constant, then the statement holds: if P(0) (a) = P(a) = 0,
then a0 = 0 so that P(x) = 0. Next, assume the statement holds for all polynomials of degree n − 1 or less,
and let P(x) be a polynomial of degree at most n with P( j) (a) = 0 for j = 0, 1, . . . , n. If P(x) has degree less
than n, then we know P(x) = 0 by induction, so assume the degree of P(x) is exactly n. Then

P(x) = an xn + an−1 xn−1 + · · · + a1 x + a0

where an  0; also,

P (x) = nan xn−1 + (n − 1)an−1 xn−2 + · · · + a1

 S E C T I O N 10.8 Taylor Series (LT SECTION 11.8) 171

Note that P( j+1) (a) = (P )( j) (a) for j = 0, 1, . . . , n − 1. But then

0 = P( j+1) (a) = (P )( j) (a) for all j = 0, 1, . . . , n − 1
Since P (x) has degree at most n − 1, it follows by induction that P (x) = 0. Thus an = an−1 = · · · = a1 = 0
so that P(x) = a0 . But P(a) = 0 so that a0 = 0 as well and thus P(x) = 0.
(b) Suppose Q(x) is a polynomial of degree at most n that agrees with f (x) at x = a up to order n. Let
P(x) = T n (x) − Q(x). Note that P(x) is a polynomial of degree at most n since both T n (x) and Q(x) are. Since
both T n (x) and Q(x) agree with f (x) at x = a to order n, we have
T n( j) (a) = f ( j) (a) = Q( j) (a), j = 0, 1, 2, . . . , n
Thus
P( j) (a) = T n( j) (a) − Q( j) (a) = 0 for j = 0, 1, 2, . . . , n
But then by part (a), P(x) = 0 so that T n (x) = Q(x).

10.8 Taylor Series (LT Section 11.8)

Preliminary Questions
1. Determine f (0) and f  (0) for a function f with Maclaurin series

T (x) = 3 + 2x + 12x2 + 5x3 + · · ·

solution The Maclaurin series for a function f has the form
f  (0) f  (0) 2 f  (0) 3
f (0) + x+ x + x + ···
1! 2! 3!
f  (0)
Matching this general expression with the given series, we find f (0) = 3 and = 5. From this latter
3!
equation, it follows that f  (0) = 30.
2. Determine f (−2) and f (4) (−2) for a function with Taylor series

T (x) = 3(x + 2) + (x + 2)2 − 4(x + 2)3 + 2(x + 2)4 + · · ·

solution The Taylor series for a function f centered at x = −2 has the form

f  (−2) f  (−2) f  (−2) f (4) (−2)

f (−2) + (x + 2) + (x + 2)2 + (x + 2)3 + (x + 2)4 + · · ·
1! 2! 3! 4!
f (4) (−2)
Matching this general expression with the given series, we find f (−2) = 0 and = 2. From this latter
4!
equation, it follows that f (4) (−2) = 48.
3. What is the easiest way to find the Maclaurin series for the function f (x) = sin(x2 )?

solution The easiest way to find the Maclaurin series for sin x2 is to substitute x2 for x in the Maclaurin
series for sin x.
4. Find the Taylor series for f centered at c = 3 if f (3) = 4 and f  (x) has a Taylor expansion
∞
(x − 3)n
f  (x) =
n=1
n

solution Integrating the series for f  (x) term-by-term gives


∞
(x − 3)n+1
f (x) = C +
n=1
n(n + 1)

Substituting x = 3 then yields

f (3) = C = 4
172 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

so

∞
(x − 3)n+1
f (x) = 4 +
n=1
n(n + 1)

5. Let T (x) be the Maclaurin series of f (x). Which of the following guarantees that f (2) = T (2)?
(a) T (x) converges for x = 2.
(b) The remainder Rk (2) approaches a limit as k → ∞.
(c) The remainder Rk (2) approaches zero as k → ∞.
solution The correct response is (c): f (2) = T (2) if and only if the remainder Rk (2) approaches zero as
k → ∞.

Exercises
1. Write out the first four terms of the Maclaurin series of f (x) if

f (0) = 2, f  (0) = 3, f  (0) = 4, f  (0) = 12

solution The first four terms of the Maclaurin series of f (x) are

f  (0) 2 f  (0) 3 4 12 3

f (0) + f  (0)x + x + x = 2 + 3x + x2 + x = 2 + 3x + 2x2 + 2x3
2! 3! 2 6

2. Write out the first four terms of the Taylor series of f (x) centered at c = 3 if

f (3) = 1, f  (3) = 2, f  (3) = 12, f  (3) = 3

solution The first four terms of the Taylor series centered at c = 3 are:

f  (3) f  (3) 12 3

f (3) + f  (3)(x − 3) + (x − 3)2 + (x − 3)3 = 1 + 2(x − 3) + (x − 3)2 + (x − 3)3
2! 3! 2 6
1
= 1 + 2(x − 3) + 6(x − 3)2 + (x − 3)3
2

In Exercises 3–18, find the Maclaurin series and find the interval on which the expansion is valid.
1
3. f (x) =
1 + 10x
solution Substituting −10x for x in the Maclaurin series for 1
1−x gives

1  ∞  ∞
= (−10x)n = (−1)n 10n xn
1 + 10x n=0 n=0

This series is valid for |10x| < 1, or |x| < 1

10 .

x2
4. f (x) =
1 − x3
solution Substituting x3 for x in the Maclaurin series for 1
1−x gives

1  ∞  ∞
= (x3 )n = x3n
1−x 3
n=0 n=0

Therefore
x2  ∞  ∞
= x2 x3n = x3n+2
1−x 3
n=0 n=0

This series is valid for |x3 | < 1, or |x| < 1.

 S E C T I O N 10.8 Taylor Series (LT SECTION 11.8) 173

5. f (x) = cos 3x
solution Substituting 3x for x in the Maclaurin series for cos x gives
∞
(3x)2n 
∞
9n x2n
cos 3x = (−1)n = (−1)n
n=0
(2n)! n=0
(2n)!
This series is valid for all x.
6. f (x) = sin(2x)
solution Substituting 2x for x in the Maclaurin series for sin x gives

∞
(2x)2n+1  ∞
22n+1 x2n+1
sin 2x = (−1)n = (−1)n
n=0
(2n + 1)! n=0 (2n + 1)!
This series is valid for all x.
7. f (x) = sin(x2 )
solution Substituting x2 for x in the Maclaurin series for sin x gives

∞
(x2 )2n+1  ∞
x4n+2
sin x2 = (−1)n = (−1)n
n=0
(2n + 1)! n=0 (2n + 1)!
This series is valid for all x.
8. f (x) = e4x
solution Substituting 4x for x in the Maclaurin series for e x gives

∞
(4x)n 
∞
4n xn
e4x = =
n=0
n! n=o
n!
This series is valid for all x.
9. f (x) = ln(1 − x2 )
solution Substituting −x2 for x in the Maclaurin series for ln(1 + x) gives

∞
(−1)n−1 (−x2 )n 
∞
(−1)2n−1 x2n ∞
x2n
ln(1 − x2 ) = = =−
n=1
n n=1
n n=1
n
This series is valid for |x| < 1.
10. f (x) = (1 − x)−1/2
solution Substituting −x for x and using a = − 12 in the Binomial series gives
∞ 
  
∞  1 
−1/2 − 12 −2
(1 − x) = (−x) =
n
(−1) n
xn
n n
n=0 n=0

This series is valid for |x| < 1.

11. f (x) = tan−1 (x2 )
solution Substituting x2 for x in the Maclaurin series for tan−1 x gives

∞
(x2 )2n+1 
∞
x4n+2
tan−1 (x2 ) = (−1)n = (−1)n
n=0
2n + 1 n=0
2n + 1
This series is valid for |x| ≤ 1.
2
12. f (x) = x2 e x
solution First substitute x2 for x in the Maclaurin series for e x to obtain

∞
(x2 )n ∞
x2n
x2
e = =
n=0
n! n=0
n!

Now, multiply by x2 to obtain

2
∞
x2n  x2n+2
∞
x2 e x = x2 =
n=0
n! n=0
n!
This series is valid for all x.
174 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

13. f (x) = e x−2

solution e x−2 = e−2 e x ; thus,

∞
xn  xn
∞
e x−2 = e−2 =
n=0
n! n=0 e2 n!

This series is valid for all x.

1 − cos x
14. f (x) =
x
∞
x2n
solution cos x = (−1)n , so
n=0
(2n)!

1 − cos x 
∞
x2n−1
= (−1)n+1
x n=1
(2n)!

This series is valid for all x.

15. f (x) = ln(1 − 5x)
solution Substituting −5x for x in the Maclaurin series for ln(1 + x) gives


∞
(−1)n−1 (−5x)n 
∞
(−1)2n−1 5n xn 
∞
5n xn
ln(1 − 5x) = = =−
n=1
n n=1
n n=1
n

This series is valid for |5x| < 1, or |x| < 15 , and for x = − 15 .
16. f (x) = (x2 + 2x)e x
solution Using the Maclaurin series for e x , we find

  ∞  
xn  xn+2  2xn+1 
∞ ∞ ∞ ∞
xn 1 2 n+1
(x + 2x)e = x
2 x
+ 2x
2
= + = 2x + + x
n=0
n! n=0
n! n=0 n! n=0
n! n=1
(n − 1)! n!

∞
n+2 
∞
n+2
= 2x + xn+1 = xn+1
n=1
n! n=0
n!

This series is valid for all x.

17. f (x) = sinh x
solution Recall that

1 x
sinh x = (e − e−x )
2
Therefore,
⎛∞ ⎞
1 ⎜⎜⎜⎜ xn  (−x)n ⎟⎟⎟⎟  xn
∞ ∞
sinh x = ⎜ − ⎟ = (1 − (−1)n )
2 ⎝ n=0 n! n=0 n! ⎠ n=0 2(n!)

Now,
⎧
⎪
⎪
⎨0, n even
1 − (−1) = ⎪
⎪
n
⎩2, n odd
so

∞
x2k+1  x2k+1
∞
sinh x = 2 =
k=0
2(2k + 1)! k=0 (2k + 1)!

This series is valid for all x.

 S E C T I O N 10.8 Taylor Series (LT SECTION 11.8) 175

18. f (x) = cosh x

solution Recall that
1 x
cosh x = (e + e−x )
2
Therefore,
⎛∞ ⎞
1 ⎜⎜⎜⎜ xn  (−x)n ⎟⎟⎟⎟  xn
∞ ∞
cosh x = ⎜⎝ + ⎟⎠ = (1 + (−1)n )
2 n=0 n! n=0 n! n=0
2(n!)

Now,
⎧
⎪
⎪
⎨0, n odd
1 + (−1) = ⎪
⎪
n
⎩2, n even
so

∞
x2k  x2k
∞
cosh x = 2 =
k=0
2(2k)! k=0 (2k)!

This series is valid for all x.

In Exercises 19–30, find the terms through degree 4 of the Maclaurin series of f (x). Use multiplication and
substitution as necessary.
19. f (x) = e x sin x

solution Multiply the fourth-order Taylor polynomials for e x and sin x:

  
x2 x3 x4 x3 x3 x3 x4 x4
1+x+ + + x− = x + x2 − + − + + higher-order terms
2 6 24 6 6 2 6 6
x3
= x + x2 + + higher-order terms
3
The terms in the Maclaurin series through degree 4 for f (x) = e x sin x are therefore
x3
x + x2 +
3
20. f (x) = e x ln(1 − x)

solution Multiply the fourth-order Taylor polynomials for e x and ln(1 − x):
  
x2 x3 x4 x2 x3 x4
1+x+ + + −x − − −
2 6 24 2 3 4
x2 x3 x3 x3 x4 x4 x4 x4
= −x − − x2 − − − − − − − + higher-order terms
2 3 2 2 4 3 4 6
3x2 4x3
= −x − − − x4 + higher-order terms
2 3
The first four terms of the Maclaurin series for f (x) = e x ln(1 − x) are therefore
3x2 4x3
−x − − − x4
2 3
sin x
21. f (x) =
1−x
1
solution Multiply the fourth order Taylor polynomials for sin x and :
1−x
 
x3  x3 x4
x− 1 + x + x2 + x3 + x4 = x + x2 − + x3 + x4 − + higher-order terms
6 6 6
5x3 5x4
= x + x2 + + + higher-order terms
6 6
176 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

sin x
The terms of the Maclaurin series through degree 4 for f (x) = are therefore
1−x

5x3 5x4
x + x2 + +
6 6

1
22. f (x) =
1 + sin x
solution Substituting sin x for x in the Maclaurin series for 1
1−x and then using the Maclaurin series for
sin x gives

1
= 1 − sin x + sin2 x − sin3 x + sin4 x − · · ·
1 + sin x
   2  3  4
x3 x3 x3 x3
=1− x− + ··· + x − + ··· − x − + ··· + x − + ··· − ···
6 6 6 6
5 3 2 4
= 1 − x + x2 − x + x + higher-order terms
6 3
1
Therefore, the terms of the Maclaurin series through degree four for f (x) = are
1 + sin x
5 3 2 4
1 − x + x2 − x + x
6 3

23. f (x) = (1 + x)1/4

solution The first five generalized binomial coefficients for a = 1

4 are
     
1 −3 1 −3 −7 1 −3 −7 −11
1 4 4 3 4 4 4 7 4 4 4 4 −77
1, , =− , = , =
4 2! 32 3! 128 4! 2048

Therefore, the first four terms in the binomial series for (1 + x)1/4 are

1 3 2 7 3 77 4
1+ x− x + x − x
4 32 128 2048

24. f (x) = (1 + x)−3/2

solution The first five generalized binomial coefficients for a = − 32 are

3 − 32 (− 25 ) 15 − 32 (− 25 )(− 27 ) 35 − 32 (− 25 )(− 27 )(− 29 ) 315

1, − , = , =− , =
2 2! 8 3! 16 4! 128

Therefore, the first five terms in the binomial series for f (x) = (1 + x)−3/2 are

3 15 2 35 3 315 4
1− x+ x − x + x
2 8 16 128

25. f (x) = e x tan−1 x

solution Using the Maclaurin series for e x and tan−1 x, we find

  
x2 x3 x3 x3 x3 x4 x4
e x tan−1 x = 1 + x + + + ··· x − + · · · = x + x2 − + + − + ···
2 6 3 3 2 6 3
1 3 1 4
= x + x2 + x − x + ···
6 6
 S E C T I O N 10.8 Taylor Series (LT SECTION 11.8) 177

26. f (x) = sin (x3 − x)

solution Substitute x3 − x into the first two terms of the Maclaurin series for sin x:

(x3 − x)3 x9 − 3x7 + 3x5 − x3

(x3 − x) − = x3 − x −
3! 3!
so that the terms of the Maclaurin series for sin(x3 − x) through degree four are
7 3
−x + x
6

27. f (x) = esin x

solution Substituting sin x for x in the Maclaurin series for e x and then using the Maclaurin series for sin x,
we find
sin2 x sin3 x sin4 x
esin x = 1 + sin x + + + + ···
2 6 24
   2
x3 1 x3 1 1
=1+ x− + ··· + x− + · · · + (x − · · · )3 + (x − · · · )4
6 2 6 6 24
1 2 1 3 1 3 1 4 1 4
=1+x+ x − x + x − x + x + ···
2 6 6 6 24
1 1 4
= 1 + x + x2 − x + ···
2 8

x
28. f (x) = e(e )
x
solution With f (x) = e(e ) , we find

f  (x) = e(e ) · e x
x


f  (x) = e(e ) · e x + e(e ) · e2x = e(e ) e2x + e x
x x x

 
f  (x) = e(e ) 2e2x + e x + e(e ) e2x + e x e x
x x

x

= e(e ) e3x + 3e2x + e x
x
 x

f (4) (x) = e(e ) 3e3x + 6e2x + e x + e(e ) e3x + 3e2x + e x e x
x

= e(e ) e4x + 6e3x + 7e2x + e x

and

f (0) = e, f  (0) = e, f  (0) = 2e, f  (0) = 5e, f (4) (0) = 15e
x
Therefore, the first terms of the Maclaurin for f (x) = e(e ) through degree four are
5e 3 15e 4
e + ex + ex2 + x + x
6 24

29. f (x) = cosh(x2 )

solution From Exercise 18, the Maclaurin series for cosh x is

∞
x2k
cosh x =
k=0
(2k)!

Substituting x2 for x gives

∞
x4k
cosh(x2 ) =
k=0
(2k)!
178 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

Thus the terms through degree 4 are

1 4
1+ x + ···
2

30. f (x) = sinh(x) cosh(x)

solution From Exercises 17 and 18, the Maclaurin series for sinh x and cosh x are


∞
x2k+1
sinh x =
k=0
(2k + 1)!
∞
x2k
cosh x =
k=0
(2k)!

So to compute terms through degree 4 for f (x), we multiply the first few terms of these two series together:
  
1 3 1 2 1 4 2
x + x + ··· 1 + x + x + · · · = x + x3 + · · ·
6 2 24 3

In Exercises 31–40, find the Taylor series centered at c and the interval on which the expansion is valid.
1
31. f (x) = , c=1
x
solution Write

1 1
=
x 1 + (x − 1)

and then substitute −(x − 1) for x in the Maclaurin series for 1

1−x to obtain

1  
∞ ∞
= [−(x − 1)]n = (−1)n (x − 1)n
x n=0 n=0

This series is valid for |x − 1| < 1.

32. f (x) = e3x , c = −1

solution Write

e3x = e3(x+1)−3 = e−3 e3(x+1)

Now, substitute 3(x + 1) for x in the Maclaurin series for e x to obtain


∞
(3(x + 1))n 
∞
3n
e3(x+1) = = (x + 1)n
n=0
n! n=0
n!

Thus,

∞
3n 
∞
3n e−3
e3x = e−3 (x + 1)n = (x + 1)n
n=0
n! n=0
n!

This series is valid for all x.

1
33. f (x) = , c=5
1−x
solution Write

1 1 1 1
= =− ·
1 − x −4 − (x − 5) 4 1 + x−5
4
 S E C T I O N 10.8 Taylor Series (LT SECTION 11.8) 179

Substituting − x−5 1
4 for x in the Maclaurin series for 1−x yields

 ∞  n  ∞
1 x−5 n (x − 5)
n
= − = (−1)
1 + x−5
4 n=0
4 n=0
4n

Thus,
1 (x − 5)n 
∞ ∞
n+1 (x − 5)
n
1
=− (−1)n = (−1)
1−x 4 n=0 4n n=0
4n+1
 
This series is valid for  x−5 
4 < 1, or |x − 5| < 4.

π
34. f (x) = sin x, c=
2
solution Note that the odd derivatives of sin x are zero at π2 , and the even derivatives alternate between +1
and −1. Thus the Taylor series centered at π2 is
 2n
∞ x − π2
n
(−1)
n=0
(2n)!

35. f (x) = x4 + 3x − 1, c=2

solution To determine the Taylor series with center c = 2, we compute

f  (x) = 4x3 + 3, f  (x) = 12x2 , f  (x) = 24x

and f (4) (x) = 24. All derivatives of order five and higher are zero. Now,
f (2) = 21, f  (2) = 35, f  (2) = 48, f  (2) = 48
and f (4) (2) = 24. Therefore, the Taylor series is
48 48 24
21 + 35(x − 2) + (x − 2)2 + (x − 2)3 + (x − 2)4
2 6 24
or
21 + 35(x − 2) + 24(x − 2)2 + 8(x − 2)3 + (x − 2)4

36. f (x) = x4 + 3x − 1, c=0

solution The function x + 3x − 1 is a polynomial in x, hence it is already in the form of a Maclaurin series.
4

1
37. f (x) = , c=4
x2
solution We will first find the Taylor series for 1
x and then differentiate to obtain the series for 1
x2
. Write
1 1 1 1
= = ·
x 4 + (x − 4) 4 1 + x−4
4

Now substitute − x−4 1

4 for x in the Maclaurin series for 1−x to obtain
∞  n 
1 1 x−4
∞
(x − 4)n
= − = (−1)n n+1
x 4 n=0 4 n=0
4

Differentiating term-by-term yields

1 
∞
(x − 4)n−1
− 2
= (−1)n n
x n=1
4n+1

so that

∞ ∞
n−1 (x − 4) (x − 4)n
n−1
1
= (−1) n = (−1) n
(n + 1)
x2 n=1 4n+1 n=0
4n+2
 
This series is valid for  x−4 
4 < 1, or |x − 4| < 4.
180 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

√
38. f (x) = x, c=4
solution Write

√  x−4
x= 4 + (x − 4) = 2 1+
4
Substituting 4for x in the binomial series with a = 12 yields
x−4

∞  1  n  ∞  1 
√ x−4 1
x=2 2 = 2 (x − 4)n
n 4 2 2n−1 n
n=0 n=0
 
This series is valid for  x−4 
4 < 1, or |x − 4| < 4.
1
39. f (x) = , c=3
1 − x2
solution By partial fraction decomposition
1 1
1
= 2 + 2
1 − x2 1 − x 1 + x
so
1 1
1 1 1 1 1
= 2
+ 2
=− · + ·
1 − x2 −2 − (x − 3) 4 + (x − 3) 4 1 + x−3
2
8 1 + x−3
4

Substituting − x−3 1
2 for x in the Maclaurin series for 1−x gives

∞  n  ∞
1 x−3 (−1)n
= − = (x − 3)n
1 + x−3
2 n=0
2 n=0
2n

while substituting − x−3

4 for x in the same series gives

∞  n  ∞
1 x−3 (−1)n
= − = (x − 3)n
1+ 4 x−3
n=0
4 n=0
4n

Thus,
1  (−1)n 1  (−1)n  
∞ ∞ ∞ ∞
1 (−1)n+1 (−1)n
= − (x − 3) n
+ (x − 3) n
= (x − 3) n
+ (x − 3)n
1 − x2 4 n=0 2n 8 n=0 4n n=0
2n+2
n=0
22n+3

∞    ∞
(−1) n+1
(−1) n
(−1) (2 − 1)
n+1 n+1
= n+2
+ 2n+3 (x − 3)n = 2n+3
(x − 3)n
n=0
2 2 n=0
2

This series is valid for |x − 3| < 2.

1
40. f (x) = , c = −1
3x − 2
solution Write
1 1 1 1
= =−
3x − 2 −5 + 3(x + 1) 5 1 − 3(x+1)
5
3(x+1) 1
and then substitute 5 for x in the Maclaurin series for to obtain
1−x

∞  n  ∞
1 3(x + 1) 3n
3(x+1)
= = n
(x + 1)n
1− 5 n=0
5 n=0
5

Thus,
1 ∞
3n
=− (x + 1)n
3x − 2 n=0
5n+1

 
This series is valid for  3(x+1)
5
 < 1, or |x + 1| < 53 .
 S E C T I O N 10.8 Taylor Series (LT SECTION 11.8) 181

41. Use the identity cos2 x = 12 (1 + cos 2x) to find the Maclaurin series for f (x) = cos2 x.
solution The Maclaurin series for cos 2x is
∞
(2x)2n 
∞
22n x2n
(−1)n = (−1)n
n=0
(2n)! n=0
(2n)!

so the Maclaurin series for cos2 x = 12 (1 + cos 2x) is

 #
1+ 1+ ∞ n 22n x2n
n=1 (−1) (2n)!

∞
22n−1 x2n
=1+ (−1)n
2 n=1
(2n)!

42. Show that for |x| < 1,

x3 x5
tanh−1 x = x + + + ···
3 5
d 1
Hint: Recall that tanh−1 x = .
dx 1 − x2
solution Because

d 1  ∞  ∞
tanh−1 x = = (x 2 n
) = x2n
dx 1 − x2 n=0 n=0

we have
∞
x2n+1 x3 x5
tanh−1 x = C + =C+x+ + + ···
n=0
2n + 1 3 5

Now, tanh−1 0 = 0, so it follows that C = 0, and

∞
x2n+1 x3 x5
tanh−1 x = = x+ + + ···
n=0
2n + 1 3 5

43. Use the Maclaurin series for ln(1 + x) and ln(1 − x) to show that
 
1 1+x x3 x5
ln = x+ + + ···
2 1−x 3 5
for |x| < 1. What can you conclude by comparing this result with that of Exercise 42?
solution Using the Maclaurin series for ln (1 + x) and ln (1 − x), we have for |x| < 1

∞
(−1)n−1 
∞
(−1)n−1
ln(1 + x) − ln(1 − x) = xn − (−x)n
n=1
n n=1
n

∞
(−1)n−1 n  xn  1 + (−1)n−1 n
∞ ∞
= x + = x
n=1
n n=1
n n=1
n

Since 1 + (−1)n−1 = 0 for even n and 1 + (−1)n−1 = 2 for odd n,


∞
2
ln (1 + x) − ln (1 − x) = x2k+1
k=0
2k + 1

Thus,
 
1 2 
∞ ∞
1 1+x 1 x2k+1
ln = (ln(1 + x) − ln(1 − x)) = x2k+1 =
2 1−x 2 2 k=0 2k + 1 k=0
2k + 1

Observe that this is the same series we found in Exercise 42; therefore,
 
1 1+x
ln = tanh−1 x
2 1−x
182 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

1 1
44. Differentiate the Maclaurin series for twice to find the Maclaurin series of .
1−x (1 − x)3
1
solution Differentiating the Maclaurin series for term-by-term, we obtain
1−x

1  ∞
= nxn−1
(1 − x) 2
n=1

Differentiating again then yields

2  ∞
= n(n − 1)xn−2
(1 − x)3 n=2

so that

1  n(n − 1)
∞  (n + 2)(n + 1)
∞
= xn−2 = xn
(1 − x) 3
n=2
2 n=0
2

1
45. Show, by integrating the Maclaurin series for f (x) = √ , that for |x| < 1,
1 − x2
∞
1 · 3 · 5 · · · (2n − 1) x2n+1
sin−1 x = x +
n=1
2 · 4 · 6 · · · (2n) 2n + 1

solution From Example 11, we know that for |x| < 1

1 
∞
1 · 3 · 5 · · · (2n − 1) 
∞
1 · 3 · 5 · · · (2n − 1)
√ = x2n = 1 + x2n
1 − x2 n=0
2 · 4 · 6 · · · (2n) n=1
2 · 4 · 6 · · · (2n)

so, for |x| < 1,

dx ∞
1 · 3 · 5 · · · (2n − 1) x2n+1
sin−1 x = √ =C+x+
1 − x2 n=1
2 · 4 · 6 · · · (2n) 2n + 1

Since sin−1 0 = 0, we find that C = 0. Thus,

∞
1 · 3 · 5 · · · (2n − 1) x2n+1
sin−1 x = x +
n=1
2 · 4 · 6 · · · (2n) 2n + 1

46. Use the first five terms of the Maclaurin series in Exercise 45 to approximate sin−1 12 . Compare the result
with the calculator value.
solution From Exercise 45 we know that for |x| < 1,

∞
1 · 3 · 5 · · · (2n − 1) x2n+1
sin−1 x = x +
n=1
2 · 4 · 6 · · · (2n) 2n + 1

The first five terms of the series are:

1 x3 1 · 3 x5 1 · 3 · 5 x7 1 · 3 · 5 · 7 x9 x3 3x5 5x7 35x9
x+ + + + = x+ + + +
2 3 2·4 5 2·4·6 7 2·4·6·8 9 6 40 112 1152
1
Setting x = , we obtain the following approximation:
2
 3  5  7  9
1 1
1
3 · 12 5 · 12 35 · 12
sin−1 ≈ +
2
+ + + ≈ 0.52358519539
2 2 6 40 112 1152
The calculator value is sin−1 12 ≈ 0.5235988775.
 S E C T I O N 10.8 Taylor Series (LT SECTION 11.8) 183

47. How many terms of the Maclaurin series of f (x) = ln(1 + x) are needed to compute ln 1.2 to within an
error of at most 0.0001? Make the computation and compare the result with the calculator value.
solution Substitute x = 0.2 into the Maclaurin series for ln (1 + x) to obtain:


∞
(0.2)n 
∞
1
ln 1.2 = (−1)n−1 = (−1)n−1 n
n=1
n n=1
5 n

1
This is an alternating series with an = . Using the Error Bound for alternating series
n · 5n
1
|ln 1.2 − S N | ≤ aN+1 =
(N + 1)5N+1
so we must choose N so that
1
< 0.0001 or (N + 1)5N+1 > 10000
(N + 1)5N+1

For N = 3, (N + 1)5N+1 = 4 · 54 = 2500 < 10, 000, and for N = 4, (N + 1)5N+1 = 5 · 55 = 15, 625 > 10, 000;
thus, the smallest acceptable value for N is N = 4. The corresponding approximation is:


4
(−1)n−1 1 1 1 1
S4 = = − + − = 0.182266666
n=1
5n · n 5 52 · 2 53 · 3 54 · 4

Now, ln 1.2 = 0.182321556, so

|ln 1.2 − S 4 | = 5.489 × 10−5 < 0.0001

48. Show that

π3 π5 π7
π− + − + ···
3! 5! 7!
converges to zero. How many terms must be computed to get within 0.01 of zero?
solution Set x = π in the Maclaurin series for sin x to obtain:

π3 π5 π7
0 = sin π = π − + − + ···
3! 5! 7!
Using the Error Bound for an alternating series, we have
 
 N
(−1)n π2n+1  π2N+3
0 −  ≤
 n=0
(2n + 1) !  (2N + 3) !

N = 4 is the smallest value for which the Error Bound is less than 0.01, so five terms are needed.
$x 2
49. Use the Maclaurin expansion for e−t to express the function F(x) = 0 e−t dt as an alternating power
2

series in x (Figure 3).

(a) How many terms of the Maclaurin series are needed to approximate the integral for x = 1 to within an
error of at most 0.001?
(b) Carry out the computation and check your answer using a computer algebra system.
solution Substituting −t2 for t in the Maclaurin series for et yields


∞
(−t2 )n ∞
t2n
e−t =
2
= (−1)n
n=0
n! n=0
n!

thus,
x 
∞
x2n+1
e−t dt =
2
(−1)n
0 n=0
n!(2n + 1)
184 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

(a) For x = 1,
1 
∞
1
e−t dt =
2
(−1)n
0 n=0
n!(2n + 1)

This is an alternating series with an = n!(2n+1)

1
; therefore, the error incurred by using S N to approximate the
value of the definite integral is bounded by
 1 
  1
 e dt − S N  ≤ aN+1 =
−t2
0  (N + 1)!(2N + 3)
To guarantee the error is at most 0.001, we must choose N so that
1
< 0.001 or (N + 1)!(2N + 3) > 1000
(N + 1)!(2N + 3)
For N = 3, (N + 1)!(2N + 3) = 4! · 9 = 216 < 1000 and for N = 4, (N + 1)!(2N + 3) = 5! · 11 = 1320 > 1000;
thus, the smallest acceptable value for N is N = 4. The corresponding approximation is

4
(−1)n 1 1 1 1
S4 = =1− + − + = 0.747486772
n=0
n!(2n + 1) 3 2! · 5 3! · 7 4! · 9

(b) Using a computer algebra system, we find

1
e−t dt = 0.746824133
2

therefore
 
 1

e−t dt − S 4  = 6.626 × 10−4 < 10−3
2
 
0

y
y = F(x)

y = T15 (x)

x
1 2
x
e−t dt.
2
FIGURE 3 The Maclaurin polynomial T 15 (x) for F(t) =
0

x
sin t dt
50. Let F(x) = . Show that
0 t
x3 x5 x7
F(x) = x − + − + ···
3 · 3! 5 · 5! 7 · 7!
Evaluate F(1) to three decimal places.
solution Divide the Maclaurin series for sin t by t to obtain

sin t 1  
∞ ∞
t2n+1 t2n
= (−1)n = (−1)n
t t n=0 (2n + 1)! n=0 (2n + 1)!

Integrating both sides of this equation and using term-by-term integration, we find
x
sin t  ∞
x2n+1 x3 x5 x7
F(x) = dt = (−1)n = x− + − + ···
0 t n=0
(2n + 1)!(2n + 1) 3 · 3! 5 · 5! 7 · 7!

For x = 1,

∞
1
F(1) = (−1)n
n=0
(2n + 1)!(2n + 1)
 S E C T I O N 10.8 Taylor Series (LT SECTION 11.8) 185

This is an alternating series with an = (2n+1)1!(2n+1) ; therefore, the error incurred by using S N to approximate
the value of the definite integral is bounded by
 1 
 sin t  1
 dt − S N  ≤ aN+1 =
0 t  (2N + 3)!(2N + 3)
To guarantee the error is at most 0.0005, we must choose N so that
1
< 0.0005 or (2N + 3)!(2N + 3) > 2000
(2N + 3)!(2N + 3)
For N = 1, (2N + 3)!(2N + 3) = 5! · 5 = 600 < 2000 and for N = 2, (2N + 3)!(2N + 3) = 7! · 7 = 35280 >
2000; thus, the smallest acceptable value for N is N = 2. The corresponding approximation is

2
(−1)n 1 1
S2 = =1− + = 0.946111111
n=0
(2n + 1)!(2n + 1) 3 · 3! 5 · 5!

In Exercises 51–54, express the definite integral as an infinite series and find its value to within an error of
at most 10−4 .
1
51. cos(x2 ) dx
0

solution Substituting x2 for x in the Maclaurin series for cos x yields


∞
(x2 )2n 
∞
x4n
cos(x2 ) = (−1)n = (−1)n
n=0
(2n)! n=0
(2n)!

therefore,
1
1 
∞
x4n+1  
∞
(−1)n
cos(x ) dx =
2
(−1) n
 =
0 n=0
(2n)!(4n + 1) 0 n=0 (2n)!(4n + 1)

This is an alternating series with an = (2n)!(4n+1)

1
; therefore, the error incurred by using S N to approximate the
value of the definite integral is bounded by
 1 
  1
 cos(x ) dx − S N  ≤ aN+1 =
2
0  (2N + 2)!(4N + 5)
To guarantee the error is at most 0.0001, we must choose N so that
1
< 0.0001 or (2N + 2)!(4N + 5) > 10000
(2N + 2)!(4N + 5)
For N = 2, (2N + 2)!(4N + 5) = 6! · 13 = 9360 < 10000 and for N = 3, (2N + 2)!(4N + 5) = 8! · 17 =
685440 > 10000; thus, the smallest acceptable value for N is N = 3. The corresponding approximation is

3
(−1)n 1 1 1
S3 = =1− + − = 0.904522792
n=0
(2n)!(4n + 1) 5 · 2! 9 · 4! 13 · 6!

1
52. tan−1 (x2 ) dx
0

solution Substituting x2 for x in the Maclaurin series for tan−1 x yields


∞
(x2 )2n+1 
∞
x4n+2
tan−1 (x2 ) = (−1)n = (−1)n
n=0
2n + 1 n=0
2n + 1

therefore,
1
1 
∞
x4n+3  
∞
(−1)n
tan−1 (x2 ) dx = (−1)n  =
0 n=0
(2n + 1)(4n + 3) 0 n=0 (2n + 1)(4n + 3)
186 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

This is an alternating series with an = (2n+1)(4n+3)

1
; therefore, the error incurred by using S N to approximate
the value of the definite integral is bounded by
 1 
  1
 tan (x ) dx − S N  ≤ aN+1 =
−1 2
0  (2N + 3)(4N + 7)
To guarantee the error is at most 0.0001, we must choose N so that
1
< 0.0001 or (2N + 3)(4N + 7) > 10000
(2N + 3)(4N + 7)
For N = 33, (2N + 3)(4N + 7) = (69)(139) = 9591 < 10000 and for N = 34, (2N + 3)(4N + 7) = (71)(143) =
10153 > 10000; thus, the smallest acceptable value for N is N = 34. The corresponding approximation is

34
(−1)n
S 34 = = 0.297953297
n=0
(2n)!(4n + 1)

1
e−x dx
3
53.
0

solution Substituting −x3 for x in the Maclaurin series for e x yields


∞
(−x3 )n 
∞
x3n
e−x =
3
= (−1)n
n=0
n! n=0
n!

therefore,
1
 x3n+1  
1 ∞ ∞
−x3 (−1)n
e dx = (−1) n
 =
0 n=0
n!(3n + 1) 0 n=0 n!(3n + 1)

This is an alternating series with an = n!(3n+1)

1
; therefore, the error incurred by using S N to approximate the
value of the definite integral is bounded by
 1 
  1
 e dx − S N  ≤ aN+1 =
−x3
0  (N + 1)!(3N + 4)
To guarantee the error is at most 0.0001, we must choose N so that
1
< 0.0001 or (N + 1)!(3N + 4) > 10000
(N + 1)!(3N + 4)
For N = 4, (N + 1)!(3N + 4) = 5! · 16 = 1920 < 10000 and for N = 5, (N + 1)!(3N + 4) = 6! · 19 = 13680 >
10000; thus, the smallest acceptable value for N is N = 5. The corresponding approximation is

5
(−1)n
S5 = = 0.807446200
n=0
n!(3n + 1)

1
dx
54. √
0 x4 + 1
solution From Example 11, we know that for |x| < 1

1 
∞
1 · 3 · 5 · · · (2n − 1) ∞
(2n)! 2n
√ = x2n = x
1 − x2 n=0
2 · 4 · 6 · · · (2n) n=0
22n (n!)2

therefore,
1 ∞
(2n)! ∞
(2n)! 4n
√ = 2n 2
(−x 2 2n
) = (−1)n 2n x
x4 + 1 n=0
2 (n!) n=0
2 (n!)2

and
1
 (2n)! x4n+1  
1 ∞ ∞
dx (2n)!
√ = (−1)n 2n 2 4n + 1 
= (−1)n 2n
0 x4 + 1 n=0
2 (n!) 0 n=0
2 (4n + 1)(n!)2
 S E C T I O N 10.8 Taylor Series (LT SECTION 11.8) 187

This is an alternating series with

(2n)!
an =
22n (4n + 1)(n!)2
therefore, the error incurred by using S N to approximate the value of the definite integral is bounded by
 1 
 dx  (2N + 2)!
 √ − S N  ≤ aN+1 = 2N+2
0 x +1
4  2 (4N + 5)((N + 1)!)2
To guarantee the error is at most 0.0001, we must choose N so that
(2N + 2)!
< 0.0001
22N+2 (4N + 5)((N + 1)!)2
For N = 124,
(2N + 2)!
= 0.0001006 > 0.0001
22N+2 (4N + 5)((N + 1)!)2
and for N = 125,
(2N + 2)!
= 0.00009943 < 0.0001
22N+2 (4N + 5)((N + 1)!)2
thus, the smallest acceptable value for N is N = 125. The corresponding approximation is

125
(2n)!
S 125 = (−1)n = 0.926987328
n=0
22n (4n + 1)(n!)2

In Exercises 55–58, express the integral as an infinite series.

x
1 − cos t
55. dt, for all x
0 t
solution The Maclaurin series for cos t is

∞
t2n  ∞
t2n
cos t = (−1)n =1+ (−1)n
n=0
(2n)! n=1
(2n)!
so
∞
t2n 
∞
t2n
1 − cos t = − (−1)n = (−1)n+1
n=1
(2n)! n=1 (2n)!

and
1 − cos t 1  
∞ ∞
t2n t2n−1
= (−1)n+1 = (−1)n+1
t t n=1 (2n)! n=1 (2n)!

Thus,
x
 t2n  
∞ ∞
x
1 − cos(t) x2n
dt = (−1) n+1
 = (−1)n+1
0 t n=1
(2n)!2n 0 n=1 (2n)!2n

x
t − sin t
56. dt, for all x
0 t
solution The Maclaurin series for sin t is

∞
t2n+1  ∞
t2n+1
sin t = (−1)n =t+ (−1)n
n=0
(2n + 1)! n=1
(2n + 1)!
so

∞
t2n+1  ∞
t2n+1
t − sin t = − (−1)n = (−1)n+1
n=1
(2n + 1)! n=1 (2n + 1)!
188 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

and
t − sin t 1  
∞ ∞
t2n+1 t2n
= (−1)n+1 = (−1)n+1
t t n=1 (2n + 1)! n=1 (2n + 1)!

Thus,
 x
x
t − sin(t) ∞
t2n+1  
∞
x2n+1
dt = (−1) n+1
 = (−1)n+1
0 t n=1
(2n + 1)!(2n + 1) 0 n=1 (2n + 1)!(2n + 1)

x
57. ln(1 + t2 ) dt, for |x| < 1
0

solution Substituting t2 for t in the Maclaurin series for ln(1 + t) yields

∞
(t2 )n 
∞
t2n
ln(1 + t2 ) = (−1)n−1 = (−1)n
n=1
n n=1
n

Thus,
x
 t2n+1  
x ∞ ∞
x2n+1
ln(1 + t ) dt =2
(−1)n  = (−1)n
0 n=1
n(2n + 1) 0 n=1 n(2n + 1)

x
dt
58. √ , for |x| < 1
0 1 − t4
solution From Example 11, we know that for |t| < 1

1 
∞
1 · 3 · 5 · · · (2n − 1) ∞
(2n)! 2n
√ = t2n = t
1 − t2 n=0
2 · 4 · 6 · · · (2n) n=0
2 (n!)2
2n

therefore,

1 ∞
(2n)! 2 2n  (2n)! 4n
∞
√ = (t ) = t
1 − t4 n=0
22n (n!)2 n=0
22n (n!)2

and
x
 (2n)! t4n+1  
x ∞ ∞
dt (2n)! x4n+1
√ =  =
1 − t4 n=0 2 (n!) 4n + 1 0 n=0 2 (n!) 4n + 1
2n 2 2n 2
0


∞
59. Which function has Maclaurin series (−1)n 2n xn ?
n=0

solution We recognize that


∞ 
∞
(−1)n 2n xn = (−2x)n
n=0 n=0

is the Maclaurin series for 1

1−x with x replaced by −2x. Therefore,

∞
1 1
(−1)n 2n xn = =
n=0
1 − (−2x) 1 + 2x

60. Which function has the following Maclaurin series?


∞
(−1)k
(x − 3)k
k=0
3k+1

For which values of x is the expansion valid?

 S E C T I O N 10.8 Taylor Series (LT SECTION 11.8) 189

solution Write the series as

 ∞  k
1 x−3
∞
(−1)k
(x − 3)k = −
k=0
3k+1 3 k=0 3

which we recognize as 1
3 times the Maclaurin series for 1
1−x with x replaced by − x−3
3 . Therefore,

∞
(−1)k 1 1 1 1
(x − 3)k = · = =
k=0
3k+1 3 1 + x−3
3
3 + x − 3 x
 
The series is valid for  x−3 
3 < 1, or |x − 3| < 3.
61. Using Maclaurin series, determine to exactly what value the following series converges:

∞
(π)2n
(−1)n
n=0
(2n)!

solution The Maclaurin series for cos x is

∞
x2n
cos x = (−1)n
n=0
(2n)!

substitute x = π to get

∞
π2n
(−1)n = cos π = −1
n=0
(2n)!

62. Using Maclaurin series, determine to exactly what value the following series converges:

∞
(ln 5)n
n=0
n!

solution The Maclaurin series for e x is

∞
xn
ex =
n=0
n!

substitute x = ln 5 to get

∞
(ln 5)n
= eln 5 = 5
n=0
n!

In Exercises 61–64, use Theorem 2 to prove that the f (x) is represented by its Maclaurin series for all x.
63. f (x) = sin(x/2) + cos(x/3)
solution With f (x) = sin(x/2) + cos(x/3), we see that
1 1
f (k) (x) =
g(x/2) + k h(x/3)
2k 3
where g(x) and h(x) are each either ± sin x or ± cos x. But then for any real number x,
 (k)   1     
 f (x) =  k g(x/2) + k h(x/3) ≤  k g(x/2) +  k h(x/3) ≤ |g(x/2)| + |h(x/3)| ≤ 2
1 1 1
2 3 2 3
So all derivatives of f are bounded by 2 everywhere. So if x is any real number, choose R > |x|. Then by
Theorem 2, the Maclaurin series for f converges to f on (−R, R) and therefore converges to f at x. So it
converges to f for all x.
64. f (x) = e−x
solution For any c, choose any R > 0 and consider the interval (c − R, c + R). For f (x) = e−x , we have
 (n)   
 f (x) = (−1)n e−x  = e−x

and on (c − R, c + R), e−x is bounded above by e−(c−R) = eR−c . Thus all derivatives of f (x) are bounded by
eR−c for any x ∈ (c − R, c + R), so by Theorem 2, f (x) is represented by its Taylor series centered at c.
190 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

65. f (x) = sinh x

solution By definition, sinh x = 12 (e x − e−x ), so if both e x and e−x are represented by their Taylor series
centered at c, then so is sinh x. But the previous exercise shows that e−x is so represented, and the text shows
that e x is.
66. f (x) = (1 + x)100
solution f (x) is a polynomial, so it is equal to its Taylor series and thus is obviously represented by its
Taylor series.
In Exercises 67–70, find the functions with the following Maclaurin series (refer to Table 2 on page 612).
x6 x9 x12
67. 1 + x3 + + + + ···
2! 3! 4!
solution We recognize

x6 x9 x12  x3n  (x3 )n

∞ ∞
1 + x3 + + + + ··· = =
2! 3! 4! n=0
n! n=0
n!

as the Maclaurin series for e x with x replaced by x3 . Therefore,

x6 x9 x12 3
1 + x3 + + + + · · · = ex
2! 3! 4!

68. 1 − 4x + 42 x2 − 43 x3 + 44 x4 − 45 x5 + · · ·
solution We recognize

∞
1 − 4x + 42 x2 − 43 x3 + 44 x5 − 45 x5 + · · · = (−4x)n
n=0

as the Maclaurin series for 1

1−x with x replaced by −4x. Therefore,
1 1
1 − 4x + 42 x2 − 43 x3 + 44 x5 − 45 x5 + · · · = =
1 − (−4x) 1 + 4x

53 x3 55 x5 57 x7
69. 1 − + − + ···
3! 5! 7!
solution Note
 
53 x3 55 x5 57 x7 53 x3 55 x5 57 x7
1− + − + · · · = 1 − 5x + 5x − + − + ···
3! 5! 7! 3! 5! 7!

∞
(5x)2n+1
= 1 − 5x + (−1)n
n=0
(2n + 1)!

The series is the Maclaurin series for sin x with x replaced by 5x, so
53 x3 55 x5 57 x7
1− + − + · · · = 1 − 5x + sin(5x)
3! 5! 7!

x12 x20 x28

70. x4 − + − + ···
3 5 7
solution We recognize

x12 x20 x28  ∞

(x4 )2n+1
x4 − + − + ··· = (−1)n
3 5 7 n=0
2n + 1

as the Maclaurin series for tan−1 x with x replaced by x4 . Therefore,

x12 x20 x28
x4 − + − + · · · = tan−1 (x4 )
3 5 7
 S E C T I O N 10.8 Taylor Series (LT SECTION 11.8) 191

In Exercises 71 and 72, let

1
f (x) =
(1 − x)(1 − 2x)
71. Find the Maclaurin series of f (x) using the identity
2 1
f (x) = −
1 − 2x 1 − x
1
solution Substituting 2x for x in the Maclaurin series for gives
1−x
1 
∞ 
∞
= (2x)n = 2n xn
1 − 2x n=0 n=0

1
which is valid for |2x| < 1, or |x| < 12 . Because the Maclaurin series for is valid for |x| < 1, the two
1−x
series together are valid for |x| < 12 . Thus, for |x| < 12 ,

1 2 1 ∞ ∞
= − =2 2n xn − xn
(1 − 2x)(1 − x) 1 − 2x 1 − x n=0 n=0


∞ 
∞ 
∞ 
= 2n+1 xn − xn = 2n+1 − 1 xn
n=0 n=0 n=0

72. Find the Taylor series for f (x) at c = 2. Hint: Rewrite the identity of Exercise 71 as
2 1
f (x) = −
−3 − 2(x − 2) −1 − (x − 2)
solution Using the given identity,
2 1 2 1 1
f (x) = − =− +
−3 − 2(x − 2) −1 − (x − 2) 3 1 + 23 (x − 2) 1 + (x − 2)

Substituting − 23 (x − 2) for x in the Maclaurin series for 1

yields
1−x


∞  n
1 n 2
= (−1) (x − 2)n
1 + 23 (x − 2) n=0 3

and substituting −(x − 2) for x in the same Maclaurin series yields

1 ∞
= (−1)n (x − 2)n
1 + (x − 2) n=0
 
The first series is valid for − 32 (x − 2) < 1, or |x − 2| < 32 , and the second series is valid for |x − 2| < 1;
therefore, the two series together are valid for |x − 2| < 1. Finally, for |x − 2| < 1,
 n ⎡  n+1 ⎤
2  
∞ ∞ ∞ ⎢⎢ ⎥⎥⎥
n 2 n⎢ 2
f (x) = − (−1) (x − 2) +n
(−1) (x − 2) =
n n
(−1) ⎢⎢⎣1 − ⎥⎥⎦ (x − 2)n
3 n=0 3 n=0 n=0
3

73. When a voltage V is applied to a series circuit consisting of a resistor R and an inductor L, the current at
time t is
V 
I(t) = 1 − e−Rt/L
R
Vt
Expand I(t) in a Maclaurin series. Show that I(t) ≈ for small t.
L
solution Substituting − RtL for t in the Maclaurin series for et gives
 n
  (−1)n  R n n  (−1)n  R n n
∞ − Rt ∞ ∞
−Rt/L L
e = = t =1+ t
n=0
n! n=0
n! L n=1
n! L
192 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

Thus,
⎛ n  n ⎟
⎞  
⎜⎜⎜ ∞
(−1) R ⎟  (−1)n+1 Rt n
∞
1−e −Rt/L
= 1 − ⎜⎜⎝1 + t ⎟⎟⎟⎠ =
n

n=1
n! L n=1
n! L

and
V  (−1)n+1  Rt n Vt V  (−1)n+1  Rt n
∞ ∞
I (t) = = +
R n=1 n! L L R n=2 n! L

If t is small, then so is Rt/L, so the terms in the series are even smaller, and we find
Vt
V(t) ≈
L
74. Use the result of Exercise 73 and your knowledge of alternating series to show that
Vt  R  Vt
1− t ≤ I(t) ≤ (for all t)
L 2L L
solution Since the series for I(t) is an alternating series, we know that the true value lies between any two
successive partial sums. Since the term for n = 2 is negative, we have
S 2 ≤ I(t) ≤ S 1 for all t
Clearly S 1 = Vt
L, and
 
Vt V −1 R2 t2 Vt VR2 t2 Vt  R 
S2 = + · 2 = − = 1 − t
L R 2! L L 2RL2 L 2L

75. Find the Maclaurin series for f (x) = cos(x3 ) and use it to determine f (6) (0).
solution The Maclaurin series for cos x is

∞
x2n
cos x = (−1)n
n=0
(2n)!
3
Substituting x for x gives
∞
x6n
cos(x3 ) = (−1)n
n=0
(2n)!

Now, the coefficient of x6 in this series is

1 1 f (6) (0)
− =− =
2! 2 6!
so
6!
f (6) (0) = − = −360
2

76. Find f (7) (0) and f (8) (0) for f (x) = tan−1 x using the Maclaurin series.
solution The Maclaurin series for f (x) = tan−1 x is:
∞
x2n+1
(−1)n
n=0
2n + 1

The coefficient of x7 in this series is

(−1)3 1 f (7) (0)
=− =
7 7 7!
so
7!
f (7) (0) = − = −6! = −720
7
The coefficient of x8 is 0, so f (8) (0) = 0.
 S E C T I O N 10.8 Taylor Series (LT SECTION 11.8) 193

20
77. Use substitution to find the first three terms of the Maclaurin series for f (x) = e x . How does the
result show that f (k) (0) = 0 for 1 ≤ k ≤ 19?

solution Substituting x20 for x in the Maclaurin series for e x yields

20

∞
(x20 )n ∞
x20n
ex = =
n=0
n! n=0
n!

the first three terms in the series are then

1 40
1 + x20 + x
2
f (k) (0)
Recall that the coefficient of xk in the Maclaurin series for f is k!
. For 1 ≤ k ≤ 19, the coefficient of xk
20
in the Maclaurin series for f (x) = e x is zero; it therefore follows that

f (k) (0)
= 0 or f (k) (0) = 0
k!
for 1 ≤ k ≤ 19.
√
78. Use the binomial series to find f (8) (0) for f (x) =
1 − x2 .
√
solution We obtain the Maclaurin series for f (x) = 1 − x2 by substituting −x2 for x in the binomial series
with a = 12 . This gives

√ ∞ 
 1
 
∞  1

n
1 − x2 = 2 −x2 = (−1)n 2 x2n
n n
n=0 n=0

The coefficient of x8 is
    
1 1
2
1
2 −1 1
2 −2 1
2 −3 15 f (8) (0)
(−1) 4 2 = =− =
4 4! 16 · 4! 8!
so
−15 · 8!
f (8) (0) = = −1575
16 · 4!

79. Does the Maclaurin series for f (x) = (1 + x)3/4 converge to f (x) at x = 2? Give numerical evidence to
support your answer.

solution The Taylor series for f (x) = (1 + x)3/4 converges to f (x) for |x| < 1; because x = 2 is not contained
on this interval, the series does not converge to f (x) at x = 2. The graph below displays

N 
 3

SN = 4 2n
n
n=0

for 0 ≤ N ≤ 14. The divergent nature of the sequence of partial sums is clear.
SN
15
10
5
0 N
2 4 6 8 10 12 14
−5
−10
−15
−20
194 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

80. Explain the steps required to verify that the Maclaurin series for f (x) = e x converges to f (x) for
all x.
solution To show that the Maclaurin series for e x converges to e x for all x, we show that for any real number
c, the Maclaurin series converges to e x on an interval containing c. To do this, it suffices to show that for any
interval I = (−R, R), the Maclaurin series for e x converges to e x on I, since each real number is contained in
some such interval. By Theorem 2, it suffices to show that there is a number K that bounds all derivatives
of e x for all numbers in the interval (−R, R). But each derivative of e x is also e x , so it suffices to show that
there is a number K that bounds e x for all x ∈ (−R, R). But e x is an increasing function, so that e x < eR for all
x ∈ (−R, R). Thus K = eR is the bound we want. Theorem 2 then assures us that the Maclaurin series for e x
converges to e x on I.
√
81. Let f (x) = 1 + x.
(a) Use a graphing calculator to compare the graph of f with the graphs of the first five Taylor polynomials
for f . What do they suggest about the interval of convergence of the Taylor series?
(b) Investigate numerically whether or not the Taylor expansion for f is valid for x = 1 and x = −1.
solution
(a) The five first terms of the Binomial series with a = 12 are
     
√ 1
1 1
2 2 − 1 1 1
2 2 − 1 2 −2 3
1 1
2
1
2 −1 1
2 −2 1
2 −3
1+x=1+ x+ x2 + x + x4 + · · ·
2 2! 3! 4!
1 1 9 45 4
= 1 + x − x2 + x3 − x + ···
2 8 4 2
Therefore, the first five Taylor polynomials are
T 0 (x) = 1;
1
T 1 (x) = 1 + x;
2
1 1 2
T 2 (x) = 1 + x − x ;
2 8
1 1 2 1 3
T 3 (x) = 1 + x − x + x ;
2 8 8
1 1 2 1 3 5 4
T 4 (x) = 1 + x − x + x − x
2 8 8 128
The figure
√ displays the graphs of these Taylor polynomials, along with the graph of the function
f (x) = 1 + x, which is shown in red.

1.5

1.0

1.5

–1.0 0.5 0.5 1.0

The graphs suggest that the interval of convergence for the Taylor series is −1 < x < 1.
N  1 
(b) Using a computer algebra system to calculate S N = 2 xn for x = 1 we find
n
n=0

S 10 = 1.409931183, S 100 = 1.414073048,

S 1000 = 1.414209104
√ N  1 

which appears to be converging to 2 as expected. At x = −1 we calculate S N = 2 · (−1)n , and find
n
n=0

S 10 = 0.176197052, S 100 = 0.056348479, S 1000 = 0.017839011

which appears to be converging to zero, though slowly.
 S E C T I O N 10.8 Taylor Series (LT SECTION 11.8) 195

82. Use the first five terms of the Maclaurin series for the elliptic integral E(k) to estimate the period T of a
1-m pendulum released at an angle θ = π4 (see Example 12).
solution The period T of a pendulum of length L released from an angle θ is

L
T =4 E(k)
g

where g ≈ 9.8 m/s2 is the acceleration due to gravity, E(k) is the elliptic integral of the first kind and
k = sin 2θ . From Example 12, we know that
∞  2
π  1 · 3 · 5 · · · (2n − 1)
E(k) = k2n
2 n=0 2 · 4 · 6 · · · (2n)

With θ = π4 ,
√
π 2− 2
k = sin =
8 2
and using the first five terms of the series for E(k), we find
 π  π ⎛⎜  2  2  2  2 ⎞
π ⎟⎟⎟⎟
⎜⎜⎜ 1 2 π 1·3 4 π 1·3·5 6 π 1·3·5·7
E sin ≈ ⎝⎜1 + sin + sin + sin + sin8 ⎟
8 2 2 8 2·4 8 2·4·6 8 2·4·6·8 8⎠
= 1.633578996
Therefore,

1
T ≈4 · 1.633578996 = 2.09 seconds
9.8

83. Use Example 12 and the approximation sin x ≈ x to show that the period T of a pendulum released at an
angle θ has the following second-order approximation:
  
L θ2
T ≈ 2π 1+
g 16

solution The period T of a pendulum of length L released from an angle θ is


L
T =4 E(k)
g

where g ≈ 9.8 m/s2 is the acceleration due to gravity, E(k) is the elliptic integral of the first kind and
k = sin 2θ . From Example 12, we know that
∞  2
π  1 · 3 · 5 · · · (2n − 1)
E(k) = k2n
2 n=0 2 · 4 · 6 · · · (2n)

Using the approximation sin x ≈ x, we have

θ θ
k = sin ≈
2 2
moreover, using the first two terms of the series for E(k), we find
⎡  2  2 ⎤  
π ⎢⎢⎢⎢ 1 θ ⎥⎥⎥⎥ π θ2
E(k) ≈ ⎣⎢1 + ⎥ = 1 +
2 2 2 ⎦ 2 16

Therefore,
   
L L θ2
T =4 E(k) ≈ 2π 1+
g g 16
196 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

In Exercises 84–87, the limits can be done using multiple L’Hôpital’s Rule steps. Power series provide an
alternative approach. In each case substitute in the Maclaurin series for the trig function or the inverse trig
function involved, simplify, and compute the limit.
x2
cos x − 1 + 2
84. lim
x→0 x4
solution Using the Maclaurin series for cos x, we find

∞
x2n x2 x4  x2n
∞
cos x = (−1)n =1− + + (−1)n
n=0
(2n)! 2 24 n=3 (2n)!

Thus,

x4 
∞
x2 x2n
cos x − 1 + = + (−1)n
2 24 n=3 (2n)!

and

1 
x2 ∞
cos x − 1 + x2n−4
2
= + (−1)n
x4 24 n=3 (2n)!

Since the power series for cos x converges everywhere, and since from the previous section, a power series
is continuous within its radius of convergence, we may compute the limit of this function by setting x = 0 in
the power series, giving

1 
x2 ∞
cos x − 1 + 02n−4 1
lim 2
= + (−1)n =
x→0 x4 24 n=3 (2n)! 24

x3
sin x − x + 6
85. lim
x→0 x5
solution Using the Maclaurin series for sin x, we find
∞
x2n+1 x3 x5 
∞
x2n+1
sin x = (−1)n = x− + + (−1)n
n=0
(2n + 1)! 6 120 n=3 (2n + 1)!

Thus,

x3 x5  ∞
x2n+1
sin x − x + = + (−1)n
6 120 n=3 (2n + 1)!

and
sin x − x + x3
1  ∞
x2n−4
6
= + (−1)n
x5 120 n=3 (2n + 1)!

Note that the radius of convergence for this series is infinite, and recall from the previous section that a
convergent power series is continuous within its radius of convergence. Thus to calculate the limit of this
power series as x → 0 it suffices to evaluate it at x = 0:
⎛ ⎞
sin x − x + x6
3
⎜⎜⎜ 1  ∞ 2n−4 ⎟
⎟⎟⎟
⎜⎜⎝ n x 1 1
lim = lim + (−1) ⎟⎠ = +0=
x→0 x5 x→0 120
n=3
(2n + 1)! 120 120

tan−1 x − x cos x − 16 x3
86. lim
x→0 x5
solution Start with the Maclaurin series for tan−1 x and cos x:
∞
x2n+1 
∞
x2n
tan−1 x = (−1)n cos x = (−1)n
n=0
2n + 1 n=0
(2n)!
 S E C T I O N 10.8 Taylor Series (LT SECTION 11.8) 197

Then

∞
x2n+1
x cos x = (−1)n
n=0
(2n)!

so that

∞  
1 1
tan−1 x − x cos x = (−1)n − x2n+1
n=0
2n + 1 (2n)!
 
19 5 
∞
1 1 1
= x3 + x + (−1)n − x2n+1
6 120 n=3
2n + 1 (2n)!

and
 
19 
∞
tan−1 x − x cos x − 16 x3 1 1
= + (−1) n
− x2n−4
x5 120 n=3 2n + 1 (2n)!

Since the radius of convergence of the series for tan−1 x is 1 and that of cos x is infinite, the radius of conver-
gence of this series is 1. Recall from the previous section that a convergent power series is continuous within
its radius of convergence. Thus to calculate the limit of this power series as x → 0 it suffices to evaluate it at
x = 0:
⎛   ⎞
tan−1 x − x cos x − 16 x3 ⎜⎜⎜ 19  ∞
1 1 ⎟⎟
2n−4 ⎟
lim = lim ⎜⎝ ⎜ + (−1) n
− x ⎟⎟⎠ = 19 + 0 = 19
x→0 x 5 x→0 120 n=3
2n + 1 (2n)! 120 120

 
sin(x2 ) cos x
87. lim − 2
x→0 x4 x
solution We start with

∞
x2n+1 ∞
x2n
sin x = (−1)n cos x = (−1)n
n=0
(2n + 1)! n=0
(2n)!

so that
sin(x2 )  
∞ ∞
x4n+2 n x
4n−2
= (−1) n
= (−1)
x4 n=0
(2n + 1)!x4 n=0 (2n + 1)!

cos x 
∞ 2n−2
nx
= (−1)
x2 n=0
(2n)!

Expanding the first few terms gives

1 
∞
sin(x2 ) n x
4n−2
= − (−1)
x4 x2 n=1 (2n + 1)!

1  2n−2 ∞
cos x 1 nx
= − + (−1)
x2 x2 2 n=2 (2n)!

so that
sin(x2 ) cos x 1  
∞ ∞
x4n−2 x2n−2
− 2 = − (−1)n − (−1)n
x 4 x 2 n=1 (2n + 1)! n=2 (2n)!

Note that all terms under the summation signs have positive powers of x. Now, the radius of convergence of
the series for both sin and cos is infinite, so the radius of convergence of this series is infinite. Recall from
the previous section that a convergent power series is continuous within its radius of convergence. Thus to
calculate the limit of this power series as x → 0 it suffices to evaluate it at x = 0:
  ⎛ ⎞
sin(x2 ) cos x ⎜⎜⎜ 1 ∞
x 4n−2 ∞
x 2n−2 ⎟
⎟⎟⎟ 1 1
lim − 2 = lim ⎜⎜⎝ − (−1) n
− (−1) n
⎟⎠ = + 0 =
x→0 x4 x x→0 2
n=1
(2n + 1)! n=2
(2n)! 2 2
198 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

88. Use Euler’s Formula to express each of the following in a + bi form.

π 5π −π
(a) e 4 i (b) 4e 3 i (c) ie 2 i
solution
√ √
(a) cos π4 + i sin π4 = 22 + 22 i
  √  √
(b) 4 cos 5π3 + i sin 5π
3 = 4 1
2 − 2 i = 2 − 2 3i
3

  
(c) i cos − π2 + i sin − π2 = i(0 − i) = 1
89. Use Euler’s Formula to express each of the following in a + bi form.
3π −π
(a) −e 4 i (b) e2πi (c) 3ie 3 i
solution
  √ √  √ √
(a) − cos 3π
4 + i sin 3π
4 = − − 22 + 2
2 i = 2
2
− 2
2 i
(b) cos 2π + i sin 2π = 1  
   √ √
(c) 3i cos − π3 + i sin − π3 = 3i 1
2 − 3
2 i = 3 3
2 + 32 i

In Exercises 90–91, use Euler’s Formula to prove that the identity holds. Note the similarity between these
relationships and the definitions of the hyperbolic sine and cosine functions.
eiz +e−iz
90. cos z = 2
solution Using Euler’s identity,

eiz + e−iz (cos z + i sin z) + (cos(−z) + i sin(−z)) (cos z + i sin z) + (cos z − i sin z) 2 cos z
= = = = cos z
2 2 2 2
eiz −e−iz
91. sin z = 2i
solution Using Euler’s identity,

eiz − e−iz (cos z + i sin z) − (cos(−z) + i sin(−z)) (cos z + i sin z) − (cos z − i sin z) 2i sin z
= = = = sin z
2i 2i 2i 2i

Further Insights and Challenges

92. In this exercise, we show that the Maclaurin expansion of f (x) = ln(1 + x) is valid for x = 1.
(a) Show that for all x  −1,

1  N
(−1)N+1 xN+1
= (−1)n xn +
1 + x n=0 1+x

(b) Integrate from 0 to 1 to obtain


N
(−1)n−1 1
xN+1 dx
ln 2 = + (−1)N+1
n=1
n 0 1+x
$1
(c) Verify that the integral on the right tends to zero as N → ∞ by showing that it is smaller than 0
xN+1 dx.
(d) Prove the formula
1 1 1
ln 2 = 1 − + − + ···
2 3 4
solution
(a) Substituting −x for x in the Maclaurin series for 1
1−x yields

1  ∞
= (−1)n xn
1 + x n=0

Now, rewrite the series as


N 
∞
(−1)n xn + (−1)n xn
n=0 n=N+1
 S E C T I O N 10.8 Taylor Series (LT SECTION 11.8) 199

and use the formula for the sum of a geometric series on the second term to obtain

1  N
(−1)N+1 xN+1
= (−1)n xn +
1 + x n=0 1+x

(b) Integrate the equation derived in part (a) from 0 to 1 to obtain

1  
xn+1 1
N 1
xN+1
ln(1 + x) = (−1)n  + (−1)N+1 dx
0 n=0
n + 1 0 0 1+x
or

N
(−1)n 1
xN+1  (−1)n−1 N+1 1
xN+1
ln 2 = + (−1)N+1 dx = + (−1)N+1 dx
n=0
n+1 0 1+x n=1
n 0 1+x

(c) For 0 < x < 1,

1 1
xN+1 xN+1
0≤ ≤ xN+1 so 0 ≤ dx ≤ xN+1 dx
1+x 0 1+x 0

Now,
1
1
xN+2  1
x N+1
dx =  = → 0 as N → ∞
0 N+2 0 N+2
Thus, by the Squeeze Theorem,
1
xN+1
lim dx = 0
N→∞ 0 1+x
(d) Taking the limit as N → ∞ of the equation derived in part (b) and using the result from part (c), we find

∞
(−1)n−1 1 1 1
ln 2 = =1− + − + ···
n=1
n 2 3 4

1 t
93. Let g(t) = − .
1+t 2 1 + t2
1
π 1
(a) Show that g(t) dt = − ln 2.
0 4 2
(b) Show that g(t) = 1 − t − t2 + t3 + t4 − t5 − t6 + · · · .
(c) Evaluate S = 1 − 12 − 13 + 14 + 15 − 16 − 17 + · · · .
solution
(a)
  1
1  π 1
g(t) dt = tan−1 t − ln(t2 + 1)  = tan−1 1 − ln 2 = − ln 2
1 1
0 2 0 2 4 2
1
(b) Start with the Taylor series for 1+t :

1  ∞
= (−1)n tn
1 + t n=0

and substitute t2 for t to get

1  ∞
= (−1)n t2n = 1 − t2 + t4 − t6 + . . .
1+t 2
n=0

so that
t  ∞
= (−1)n t2n+1 = t − t3 + t5 − t7 + . . .
1 + t2 n=0

Finally,
1 t
g(t) = − = 1 − t − t2 + t3 + t4 − t5 − t6 + t7 + . . .
1 + t2 1 + t2
200 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

(c) We have
1 1 1 1 1
g(t) dt = (1 − t − t2 + t3 + t4 − t5 − . . . ) dt = t − t2 − t3 + t4 + t5 − t6 − · · · + C
2 3 4 5 6
The radius of convergence of the series for g(t) is 1, so the radius of convergence of this series is also 1.
However, this series converges at the right endpoint, t = 1, since
     
1 1 1 1 1
1− − − + − − ...
2 3 4 5 6
is an alternating series with general term decreasing to zero. Thus by part (a),
1 1 1 1 1 π 1
1− − + + − − · · · = − ln 2
2 3 4 5 6 4 2

In Exercises 94 and 95, we investigate the convergence of the binomial series

∞  
a n
T a (x) = x
n=0
n

94. Prove that T a (x) has radius of convergence R = 1 if a is not a whole number. What is the radius of
convergence if a is a whole number?
solution Suppose that a is not a whole number. Then
 
a a (a − 1) · · · (a − n + 1)
=
n n!
is never zero. Moreover,
   
 a 
 n + 1   a(a − 1) · · · (a − n + 1)(a − n)   
    =  ·
n!  =  a − n 
 a   (n + 1)! a(a − 1) · · · (a − n + 1)   n + 1 
 
n
so, by the formula for the radius of convergence
 
 a − n 
r = lim  =1
n→∞ n + 1 

−1
 istherefore R = r = 1.
The radius of convergence of T a (x)
a
If a is a whole number, then = 0 for all n > a. The infinite series then reduces to a polynomial of
n
degree a, so it converges for all x (i.e., R = ∞).
95. By Exercise 94, T a (x) converges for |x| < 1, but we do not yet know whether T a (x) = (1 + x)a .
(a) Verify the identity
     
a a a
a = n + (n + 1)
n n n+1
(b) Use (a) to show that y = T a (x) satisfies the differential equation (1 + x)y = ay with initial condition
y(0) = 1.
T a (x)
(c) Prove T a (x) = (1 + x)a for |x| < 1 by showing that the derivative of the ratio is zero.
(1 + x)a
solution
(a)
   
a a a (a − 1) · · · (a − n + 1) a (a − 1) · · · (a − n + 1) (a − n)
n + (n + 1) =n· + (n + 1) ·
n n+1 n! (n + 1) !
a (a − 1) · · · (a − n + 1) a (a − 1) · · · (a − n + 1) (a − n)
= +
(n − 1) ! n!
 
a (a − 1) · · · (a − n + 1) (n + (a − n)) a
= =a·
n! n
 S E C T I O N 10.8 Taylor Series (LT SECTION 11.8) 201

(b) Differentiating T a (x) term-by-term yields


∞  
a
T a (x) = n xn−1
n
n=1

Thus,

∞   
∞   
∞   
∞  
a a a a
(1 + x)T a (x) = n xn−1 + n xn = (n + 1) xn + n xn
n n n+1 n
n=1 n=1 n=0 n=0
∞ %
    & ∞ 
 
a a a
= (n + 1) +n xn = a xn = aT a (x)
n+1 n n
n=0 n=0

Moreover,
 
a
T a (0) = =1
0
(c)
 
d T a (x) (1 + x)a T a (x) − a(1 + x)a−1 T a (x) (1 + x)T a (x) − aT a (x)
= = =0
dx (1 + x)a (1 + x)2a (1 + x)a+1
Thus,
T a (x)
=C
(1 + x)a
for some constant C. For x = 0,
T a (0) 1
a = = 1, so C = 1
(1 + 0) 1
Finally, T a (x) = (1 + x)a .
π/2 
96. The function G(k) = 1 − k2 sin2 t dt is called an elliptic integral of the second kind. Prove that
0
for |k| < 1,
∞  2 2n
π π  1 · 3 · · · (2n − 1) k
G(k) = −
2 2 n=1 2 · · · 4 · (2n) 2n − 1

solution For |k| < 1, |k2 sin2 t| < 1 for all t. Substituting −k2 sin2 t for t in the binomial series for a = 12 , we
find
 ∞ 
 1

n
1 − k sin t = 1 +
2 2 2 −k2 sin2 t
n
n=1
  

∞ 1
2
1
2 −1 1
2 − 2 ··· 1
2 −n+1
=1+ (−1)n k2n sin2n t
n=1
n!
∞
1(1 − 2)(1 − 4) · · · (1 − 2(n − 1)) 2n 2n
=1+ (−1)n k sin t
n=1
2n n!
∞
(2 − 1)(4 − 1) · · · (2n − 3) 2n 2n
=1+ (−1)n (−1)n−1 k sin t
n=1
2n n!

∞
1 · 3 · 5 · · · (2n − 3)
=1− k2n sin2n t
n=1
2 · 4 · 6 · · · (2n)
π
Integrating from 0 to 2 term-by-term, we obtain

π  1 · 3 · 5 · · · (2n − 3) 2n
∞ π/2
G(k) = − k sin2n t dt
2 n=1 2 · 4 · 6 · · · (2n) 0
202 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

Finally, using the formula

π/2
1 · 3 · 5 · · · (2n − 1) π
sin2n t dt =
0 2 · 4 · 6 · · · (2n) 2
we arrive at
∞  2 ∞  2 2n
π π  1 · 3 · 5 · · · (2n − 3) π π  1 · 3 · 5 · · · (2n − 1) k
G(k) = − (2n − 1)k = −
2n
2 2 n=1 2 · 4 · 6 · · · (2n) 2 2 n=1 2 · 4 · 6 · · · (2n) 2n − 1

 2  2
97. Assume that a < b and let L be the arc length (circumference) of the ellipse ax + by = 1 shown in
Figure 4.There is no explicit formula for L, but it is known that L = 4bG(k), with G(k) as in Exercise 96
and k = 1 − a2 /b2 . Use the first three terms of the expansion of Exercise 96 to estimate L when a = 4 and
b = 5.
y
b

x
a

 x 2  y 2
FIGURE 4 The ellipse + = 1.
a b

solution With a = 4 and b = 5,

42 3
k= 1− =
52 5
 x 2  y 2
and the arc length of the ellipse + = 1 is
4 5
⎛ ⎞
  ⎜⎜⎜ ∞  2  3 2n ⎟⎟
3 ⎜ π π 1 · 3 · · · (2n − 1) ⎟⎟⎟
= 20 ⎜⎜⎜⎜ −
5
L = 20G ⎟⎟
5 ⎝ 2 2 n=1 2 · 4 · · · (2n) 2n − 1 ⎟⎠

Using the first three terms in the series for G(k) gives
⎛  2  2 ⎞  
⎜⎜⎜ 1 (3/5) 2
1 · 3 (3/5) 4⎟
⎟⎟⎟ 9 243 36157π
L ≈ 10π − 10π ⎜⎜⎝ · + · ⎟⎠ = 10π 1 − − = ≈ 28.398
2 1 2·4 3 100 40000 4000

98. Use Exercise 96 to prove that if a < b and a/b is near 1 (a nearly circular ellipse), then

π a2
L≈ 3b +
2 b
Hint: Use the first two terms of the series for G(k).
solution From the previous exercise, we know that

a2
L = 4bG(k), where k= 1−
b2
Following the hint and using only the first two terms of the series expansion for G(k) from Exercise 96, we
find
⎛   ⎞     
⎜⎜⎜ π π 1 2 2 ⎟⎟⎟ π a2 π a2
⎜
L ≈ 4b ⎝⎜ − ⎟
k ⎠⎟ = 4b − b 1 − 2 = 3b +
2 2 2 2 b 2 b
 Chapter Review Exercises 203

99. Irrationality of e Prove that e is an irrational number using the following argument by contradiction.
Suppose that e = M/N, where M, N are nonzero integers.
(a) Show that M! e−1 is a whole number.
(b) Use the power series for f (x) = e x at x = −1 to show that there is an integer B such that M! e−1 equals
 
1 1
B + (−1) M+1 − + ···
M + 1 (M + 1)(M + 2)
(c) Use your knowledge of alternating series with decreasing terms to conclude that 0 < |M! e−1 − B| < 1
and observe that this contradicts (a). Hence, e is not equal to M/N.
solution Suppose that e = M/N, where M, N are nonzero integers.
(a) With e = M/N,
N
M!e−1 = M! = (M − 1)!N
M
which is a whole number.
(b) Substituting x = −1 into the Maclaurin series for e x and multiplying the resulting series by M! yields
 
1 1 (−1)k
M!e−1 = M! 1 − 1 + − + ··· + + ···
2! 3! k!
M!
For all k ≤ M, is a whole number, so
k!
 
1 1 (−1)k
M! 1 − 1 + − + ··· +
2! 3! M!
is an integer. Denote this integer by B. Thus,
   
−1 (−1) M+1 (−1) M+2 1 1
M! e = B + M! + + · · · = B + (−1) M+1
− + ···
(M + 1)! (M + 2)! M + 1 (M + 1)(M + 2)

(c) The series for M! e−1 obtained in part (b) is an alternating series with an = Mn!! . Using the Error Bound
for an alternating series and noting that B = S M , we have
 
 M! e−1 − B ≤ a M+1 =
1
<1
M+1
This inequality implies that M! e−1 − B is not a whole number; however, B is a whole number so M! e−1
cannot be a whole number. We get a contradiction to the result in part (a), which proves that the original
assumption that e is a rational number is false.
100. Use the result of Exercise 75 in Section 4.5 to show that the Maclaurin series of the function
⎧
⎪
⎪
⎨e−1/x for x  0
2

f (x) = ⎪
⎪
⎩0 for x = 0
is T (x) = 0. This provides an example of a function f whose Maclaurin series converges but does not
converge to f (x) (except at x = 0).
solution By the referenced exercise, f (x) has continuous derivatives of all orders at 0, and f (n) (0) = 0 for
all n > 0. But then the Maclaurin series is
∞
f (n) (0) n ∞
f (n) (0) n
f (x) = x = f (0) + x =0
n=0
n! n=1
n!

CHAPTER REVIEW EXERCISES

n−3
1. Let an = and bn = an+3 . Calculate the first three terms in each sequence.
n!
(a) a2n (b) bn

(c) an bn (d) 2an+1 − 3an

204 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

solution
(a)
 2
1−3
a21 = = (−2)2 = 4;
1!
 2  2
2−3 1 1
a22 = = − = ;
2! 2 4
 2
3−3
a23 = =0
3!

(b)

4−3 1
b1 = a4 = = ;
4! 24
5−3 1
b2 = a5 = = ;
5! 60
6−3 1
b3 = a6 = =
6! 240

(c) Using the formula for an and the values in (b) we obtain:

1−3 1 1
a1 b1 = · =− ;
1! 24 12
2−3 1 1
a2 b2 = · =− ;
2! 60 120
3−3 1
a3 b3 = · =0
3! 240

(d)

1
2a2 − 3a1 = 2 − − 3(−2) = 5;
2
 
1 3
2a3 − 3a2 = 2 · 0 − 3 − = ;
2 2
1 1
2a4 − 3a3 = 2 · −3·0=
24 12

2n − 1 2
2. Prove that lim = using the limit definition.
n→∞ 3n + 2 3
solution Note
     
 2n − 1 − 2  =  6n − 3 − 2(3n + 2)  = − 7  = 7
<
7
 3n + 2 3   3(3n + 2)   3(3n + 2)  3(3n + 2) 9n
 
Therefore, to have an − 23  < , we need

7 7
< or n>
9n 9

Thus, let  > 0 and take M = 7

9 . Then, whenever n > M,
 
 2n − 1 − 2  = 7 7 7 9
 3n + 2 3  3(3n + 2) < 9n < 9 · 7 = 
 Chapter Review Exercises 205

In Exercises 3–8, compute the limit (or state that it does not exist) assuming that lim an = 2.
n→∞

3. lim (5an − 2a2n )

n→∞
solution
  2
lim 5an − 2a2n = 5 lim an − 2 lim a2n = 5 lim an − 2 lim an = 5 · 2 − 2 · 22 = 2
n→∞ n→∞ n→∞ n→∞ n→∞

1
4. lim
n→∞ an
1 1 1
solution lim = =
n→∞ an limn→∞ an 2
5. lim ean
n→∞
solution The function f (x) = e x is continuous, hence:

lim ean = elimn→∞ an = e2

n→∞

6. lim cos(πan )
n→∞
solution The function f (x) = cos(πx) is continuous, hence:
 
lim cos(πan ) = cos π lim an = cos(2π) = 1
n→∞ n→∞

7. lim (−1)n an
n→∞
solution Because lim an  0, it follows that lim (−1)n an does not exist.
n→∞ n→∞
an + n
8. lim
n→∞ an + n2

solution Because the sequence {an } converges, {an } is bounded and

an
lim =0
n→∞ n2
Thus,
an + n
an
n2
+ 1
n 0+0
lim = lim = =0
n→∞ an + n2 n→∞
an
n2
+1 0+1

In Exercises 9–22, determine the limit of the sequence or show that the sequence diverges.
√ √
9. an = n + 5 − n + 2
solution First rewrite an as follows:
√ √ √ √
n+5− n+2 n+5+ n+2 (n + 5) − (n + 2) 3
an = √ √ = √ √ = √ √
n+5+ n+2 n+5+ n+2 n+5+ n+2
Thus,
3
lim an = lim √ √ =0
n→∞ n→∞ n+5+ n+2

3n3 − n
10. an =
1 − 2n3
3n3 − n 3
solution lim an = lim =−
n→∞ n→∞ 1 − 2n3 2
1/n2
11. an = 2
solution The function f (x) = 2 x is continuous, so
2 2
lim an = lim 21/n = 2limn→∞ (1/n ) = 20 = 1
n→∞ n→∞
206 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

10n
12. an =
n!
solution For n > 10, write an as
      
10 10 10 10 10 10 1010 10 1010
0 ≤ an = · · ··· · · · ··· · < · =
1 2 10 11 12 n 10! n 9!n

1010 each factor is less than 1
equals
10 !
Thus, by the Squeeze Theorem, lim an = 0.
n→∞

13. bm = 1 + (−1) m

solution Because 1 + (−1)m is equal to 0 for m odd and is equal to 2 for m even, the sequence {bm } does
not approach one limit; hence this sequence diverges.
1 + (−1)m
14. bm =
m
solution The numerator is equal to zero for m odd and is equal to 2 for m even. Therefore,
1 + (−1)m 2
0≤ ≤
m m
and by the Squeeze Theorem, lim bm = 0.
m→∞
 
n+2
15. bn = tan−1
n+5
solution The function tan−1 x is continuous, so
   
n+2 n+2 π
lim bn = lim tan−1 = tan−1 lim = tan−1 1 =
n→∞ n→∞ n+5 n→∞ n + 5 4

100n 3 + πn
16. an = −
n! 5n
solution For n > 100,
 
100n 100 100 100 100 100 100 100100
0≤ = · ··· · · <
n! 1 2 100 101 102 n 99!n
therefore,
100n
lim =0
n→∞ n!

by the Squeeze Theorem. Moreover,

   π n
3 + πn 3
lim = lim + lim =0+0=0
n→∞ 5n n→∞ 5n n→∞ 5

Thus,
lim an = 0 + 0 = 0
n→∞

√ √
17. bn = n2 + n − n2 + 1
solution Rewrite bn as
√ √ √ √  
n2 + n − n2 + 1 n2 + n + n2 + 1 n2 + n − n2 + 1 n−1
bn = √ √ = √ √ = √ √
n2 + n + n2 + 1 n2 + n + n2 + 1 n2 + n + n2 + 1
Then
n
− 1
1− 1
1−0 1
lim bn = lim n n
= lim n
= √ √ =
n→∞ n→∞ n2
+ n
+ n2
+ 1 n→∞
1+ 1
+ 1+ 1 1+0+ 1+0 2
n2 n2 n2 n2 n n2
 Chapter Review Exercises 207

√ √
18. cn = n2 + n − n2 − n
solution Rewrite cn as
√ √ √ √  
n2 + n − n2 − n n2 + n + n2 − n n2 + n − n2 − n 2n
cn = √ √ = √ √ = √ √
n +n+ n −n
2 2 n +n+ n −n
2 2 n + n + n2 − n
2

Then
2n
2 2
lim cn = lim n
= lim = √ √ =1
n→∞ n→∞ n2
+ n
+ n2
− n n→∞
1+ 1
+ 1− 1 1+0+ 1−0
n2 n2 n2 n2 n n

 3m
1
19. bm = 1 +
m
 m
1
solution lim bm = lim 1 + =e
m→∞ m→∞ m
 n
3
20. cn = 1 +
n
solution Write
 n ⎡ n/3 ⎤3
1 ⎢⎢⎢ 1 ⎥⎥⎥
cn = 1 + = ⎢⎢⎣ 1 + ⎥⎥⎦
n/3 n/3

Then, because x3 is a continuous function,

⎡  n/3 ⎤3
⎢⎢⎢ 1 ⎥⎥⎥
⎢
lim cn = ⎢⎣ lim 1 + ⎥⎥⎦ = e3
n→∞ n→∞ n/3

21. bn = n ln(n + 1) − ln n
solution Write
 ln 1 +
 1
n+1 n
bn = n ln = 1
n n

Using L’Hôpital’s Rule, we find

     −1
1 −1
ln 1 + 1n ln 1 + 1
x 1+ x · − x12 1
lim bn = lim = lim = lim = lim 1 + =1
n→∞ n→∞ 1
n
1 x→∞
x
x→∞ − x2
1 x→∞ x

ln(n2 + 1)
22. cn =
ln(n3 + 1)
solution Using L’Hôpital’s Rule, we find

ln(n2 + 1) 2n/(n2 + 1) 2n4 + 2n 2 + 2n−3 2

lim cn = lim = lim 2 3 = lim 4 = lim =
n→∞ n→∞ ln(n + 1)
3 n→∞ 3n /(n + 1) n→∞ 3n + 3n 2 n→∞ 3 + 3n−2 3

arctan(n2 )
23. Use the Squeeze Theorem to show that lim √ = 0.
n→∞ n
solution For all x,
π π
− < arctan x <
2 2
so
π/2 arctan(n2 ) π/2
−√ < √ < √
n n n
208 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

for all n. Because

 
π/2 π/2
lim − √ = lim √ = 0
n→∞ n n→∞ n
it follows by the Squeeze Theorem that
arctan(n2 )
lim √ =0
n→∞ n

24. Give an example of a divergent sequence {an } such that {sin an } is convergent.
solution Let an = (−1)n π. This is an alternating series, which does not approach 0, hence it diverges.
However, an is a multiple of π for every n, and thus, sin an = 0. Since {sin an } is a constant sequence, it
converges.
an+1 1 1
25. Calculate lim , where an = 3n − 2n .
n→∞ an 2 3
solution
(a) Because
1 n 1 n 1 n 1 n 3n
3 − 2 ≥ 3 − 3 =
2 3 2 3 6
and
3n
lim =∞
n→∞ 6

we conclude that
lim an = ∞
n→∞
 n+1
an+1
1 n+1
3 − 1 n+1
2 3n+2
− 2n+2 3 − 2 2
3 3−0
(b) lim = lim 2 1 3
= lim = lim  = =3
n→∞ an n→∞
23 − 32
n 1 n n→∞ 3n+1 −2 n+1 n→∞
1− 2
n+1 1−0
3

26. Define an+1 = an + 6 with a1 = 2.
(a) Compute an for n = 2, 3, 4, 5.
(b) Show that {an } is increasing and is bounded by 3.
(c) Prove that lim an exists and find its value.
n→∞

solution
(a) We compute the first four values of an recursively:
 √ √ √
a2 = a1 + 6 = 2 + 6 = 8 = 2 2 ≈ 2.828427
 √
a3 = a2 + 6 = 2 2 + 6 ≈ 2.971267

 √
a4 = a3 + 6 = 2 2 + 6 + 6 ≈ 2.995207

 √
a5 = a4 + 6 = 2 2 + 6 + 6 + 6 ≈ 2.999201

(b) By part (a) and the given data, a2 ≈ 2.8 and a1 = 2, so a2 > a1 . Now, suppose that ak > ak−1 ; then
 
ak+1 = ak + 6 > ak−1 + 6 = ak
Thus, by mathematical induction, an+1 > an for all n and {an } is increasing.
Next, note that a1 = 2 < 3. Suppose ak < 3, then
 √
ak+1 = ak + 6 < 3 + 6 = 3
Thus, by mathematical induction, an < 3 for all n.
 Chapter Review Exercises 209

(c) Since {an } is increasing and has an upper bound, {an } converges. Let
L = lim an
n→∞

Then,
√
L= L+6
L2 = L + 6
L2 − L − 6 = 0
(L − 3)(L + 2) = 0
so L = 3 or L = −2; however, the sequence is increasing and its first term is positive, so −2 cannot be the
limit. Therefore,
lim an = 3
n→∞

∞
n−2
27. Calculate the partial sums S 4 and S 7 of the series 2 + 2n
n=1
n
solution
1 1 2 11
S4 = − + 0 + + =− = −0.183333
3 15 24 60
1 1 2 3 4 5 287
S7 = − + 0 + + + + + = = 0.065079
3 15 24 35 48 63 4410
1 1 1
28. Find the sum 1 − + − + ··· .
4 42 43
solution This is a geometric series with r = − 14 . Therefore,
1 1 1 1 4
1− + − + ··· = =
4 42 43 1 − (− 4 )
1 5

4 8 16 32
29. Find the sum + + + + ··· .
9 27 81 243
solution This is a geometric series with common ratio r = 23 . Therefore,
4
4 8 16 32 4
+ + + + ··· = 9 =
9 27 81 243 1− 2
3
3

30. Use series to determine a reduced fraction that has decimal expansion 0.121212 · · · .
solution The decimal may be regarded as a geometric series:

12 12 12  12 ∞
0.121212 . . . = + + ... =
100 10000 1000000 n=1
102n
12 1
The series has first term 100 and ratio 100 , so its sum is
12/100 12 100 12 4
0.121212 . . . = = · = =
1 − 1/100 100 99 99 33
31. Use series to determine a reduced fraction that has decimal expansion 0.108108108 · · · .
solution The decimal may be regarded as a geometric series:

108 108 108  108 ∞

0.108108108 . . . = + + . . . =
103 106 109 n=1
103n
108 1
The series has first term 103
and ratio 103
, so its sum is
108/103 108 103 108 4
0.108108108 . . . = = · = =
1 − 1/103 103 999 999 37
210 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

∞  n
 2
32. Find the sum .
n=2
e

solution This is a geometric series with common ratio r = 2e . Therefore,


∞  n
 2 2 4
2 e e2 4
= = =
n=2
e 1 − 2e 1− 2
e
e2 − 2e

∞
2n+3
33. Find the sum .
n=−1
3n

solution Note

∞ 
∞ ∞  n

2n+1 2n 2
=2 =2
n=0
3n n=0
3n n=0
3

therefore,

∞
2n+1 1
=2 =2·3=6
n=0
3n 1− 2
3


∞
π
34. Show that b − tan−1 n2 diverges if b  .
n=1
2

solution Note
 π
lim b − tan−1 n2 = b − lim tan−1 n2 = b −
n→∞ n→∞ 2
If b  π2 , then the limit of the terms in the series is not 0; hence, the series diverges by the Divergence Test.

∞ 
∞ 
∞
35. Give an example of divergent series an and bn such that (an + bn ) = 1.
n=1 n=1 n=1

1 n
solution Let an = 2 + 1, bn = −1. The corresponding series diverge by the Divergence Test; however,

∞ ∞  n 1
1
(an + bn ) = = 2 =1
n=1 n=1
2 1− 1
2

∞ 
 
1 1
36. Let S = − . Compute S N for N = 1, 2, 3, 4. Find S by showing that
n=1
n n+2

3 1 1
SN = − −
2 N+1 N+2
solution
1 2
S1 = 1 − = ;
3 3
   
1 1 1 3 7 11
S2 = 1 − + − = − = ;
3 2 4 2 12 12
     
1 1 1 1 1 3 9 21
S3 = 1 − + − + − = − = ;
3 2 4 3 5 2 20 20
       
1 1 1 1 1 1 1 3 11 17
S4 = 1 − + − + − + − = − =
3 2 4 3 5 4 6 2 30 15
The general term in the sequence of partial sums is
           
1 1 1 1 1 1 1 1 1 1 1
SN = 1 − + − + − + − + ··· + − + −
3 2 4 3 5 4 6 N−1 N+1 N N+2
 Chapter Review Exercises 211
 
1 1 1 3 1 1
=1+ − − = − +
2 N+1 N+2 2 N+1 N+2

Finally,
% &
3 1 1 3
S = lim S N = lim − + =
N→∞ N→∞ 2 N+1 N+2 2


∞
1
37. Evaluate S = .
n=3
n(n + 3)

solution Note that

 
1 1 1 1
= −
n(n + 3) 3 n n + 3

so that
 N  
1 1
N
1 1
= −
n=3
n(n + 3) 3 n=3 n n + 3
     
1 1 1 1 1 1 1
= − + − + −
3 3 6 4 7 5 8
     
1 1 1 1 1 1
− + ··· + − + −
6 9 N−1 N+2 N N+3
 
1 1 1 1 1 1 1
= + + − − −
3 3 4 5 N+1 N+2 N+3

Thus

∞ N  
1 1 1 1
= lim −
n=3
n(n + 3) 3 N→∞ n=3 n n + 3
   
1 1 1 1 1 1 1 1 1 1 1 47
= + + − − − = + + =
3 3 4 5 N+1 N+2 N+3 3 3 4 5 180

38. Find the total area of the infinitely many circles on the interval [0, 1] in Figure 1.

x
0
1 1 1 1
8 4 2

FIGURE 1

solution The diameter of the largest circle is 12 , and the diameter of each smaller circle is 1
2 the diameter of
the previous circle; thus, the diameter of the nth circle (for n ≥ 1) is 21n and the area is
 2
1 π
π =
2n+1 4n+1

The total area of the circles is

 ∞  n
π 1
∞ 1
π π π
= = · 4
=
n=1
4n+1 4 n=1
4 4 1− 1
4
12
212 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

In Exercises 39–42, use the Integral Test to determine whether the infinite series converges.

∞
n2
39.
n=1
n3 +1
x2
solution Let f (x) = x3 +1
. This function is continuous and positive for x ≥ 1. Because

(x3 + 1)(2x) − x2 (3x2 ) x(2 − x3 )

f  (x) = = 3
(x3 + 1)2 (x + 1)2
we see that f  (x) < 0 and f is decreasing on the interval x ≥ 2. Therefore, the Integral Test applies on the
interval x ≥ 2. Now,
∞
x2 R
x2 1 
dx = lim dx = lim ln(R3 + 1) − ln 9 = ∞
2 x3 + 1 R→∞ 2 x +1
3 3 R→∞


∞
n2  n2 ∞
The integral diverges; hence, the series diverges, as does the series .
n=2
n +1
3
n=1
n3 + 1

∞
n2
40.
n=1
(n3 + 1)1.01
x2
solution Let f (x) = (x3 +1)1.01
. This function is continuous and positive for x ≥ 1. Because

(x3 + 1)1.01 (2x) − x2 · 1.01(x3 + 1)0.01 (3x2 ) x(x3 + 1)0.01 (2 − 1.03x3 )

f  (x) = =
(x3 + 1)2.02 (x3 + 1)2.02
we see that f  (x) < 0 and f is decreasing on the interval x ≥ 2. Therefore, the Integral Test applies on the
interval x ≥ 2. Now,
∞ R  
x2 x2 1 1 1 1
dx = lim dx = − lim − 0.01 =
2 (x + 1) R→∞ 2 (x + 1) 0.03 R→∞ (R + 1) 0.03 · 90.01
3 1.01 3 1.01 3 0.01 9

∞
n2  ∞
n2
The integral converges; hence, the series converges, as does the series .
n=2
(n3 + 1)1.01 n=1
(n3 + 1)1.01

∞
1
41.
n=1
(n + 2)(ln(n + 2))3

solution Let f (x) = 1

(x+2) ln3 (x+2)
. Using the substitution u = ln(x + 2), so that du = 1
x+2 dx, we have
  
∞ ∞
1 ∞
1 1 R
f (x) dx = du = lim du = lim − 
0 ln 2 u
3 R→∞ ln 2 u3 R→∞ 2u2 ln 2
 
1 1 1
= lim − =
R→∞ 2(ln 2)2 2(ln R)2 2(ln 2)2
Since the integral of f (x) converges, so does the series.
∞
n3
42.
n=1
en4

Let f (x) = x3 e−x . This function is continuous and positive for x ≥ 1. Because
4
solution
 
f  (x) = x3 −4x3 e−x + 3x2 e−x = x2 e−x 3 − 4x4
4 4 4

we see that f  (x) < 0 and f is decreasing on the interval x ≥ 1. Therefore, the Integral Test applies on the
interval x ≥ 1. Now,
∞ R
1  4 1
x3 e−x dx = lim x3 e−x dx = − lim e−R − e−1 =
4 4

1 R→∞ 1 4 R→∞ 4e

∞
n3
The integral converges; hence, the series also converges.
n=1
en4
 Chapter Review Exercises 213

In Exercises 43–50, use the Direct Comparison or Limit Comparison Test to determine whether the infinite
series converges.

∞
1
43.
n=1
(n + 1)2
solution For all n ≥ 1,
1 1 1 1
0< < so <
n+1 n (n + 1)2 n2
∞
1 ∞
1
The series is a convergent p-series, so the series converges by the Direct Comparison
n=1
n2
n=1
(n + 1)2
Test.
 ∞
1
44. √
n=1
n +n
solution Apply the Limit Comparison Test with an = √1
n+n
and bn = 1n . Now,

√1
n+n n 1
L= lim 1 = lim √ = lim =1
n→∞
n
n→∞ n + n n→∞ √1
n
+1

∞
1
Because L > 0 and is the divergent harmonic series, we conclude by the Limit Comparison Test that
n=1
n

∞
1
the series √ also diverges.
n=1
n+n
∞
n2 + 1
45.
n=2
n3.5 − 2
n2 +1
solution Apply the Limit Comparison Test with an = n3.5 −2
and bn = 1
n1.5
. Now,
n2 +1
n3.5 −2 n3.5 + n1.5
L = lim = lim =1
n→∞ 1 n→∞ n3.5 − 2
n1.5
∞
1
Because L exists and 1.5
is a convergent p-series, we conclude by the Limit Comparison Test that the
n=1
n
∞
n2 + 1
series also converges.
n=2
n3.5 − 2

∞
1
46.
n=1
n − ln n
solution Since 0 ≤ ln n ≤ n for all n ≥ 1, we have 0 ≤ n − ln n ≤ n and
1 1
≤
n n − ln n

∞
1 
∞
1
The harmonic series diverges, so we conclude by the Direct Comparison Test that also
n=1
n n=1
n − ln n
diverges.
 ∞
n
47. √
n=2 n +5
5

solution For all n ≥ 2,

n n 1
√ < =
n5 +2 n5/2 n3/2

∞
1 
∞
n
The series is a convergent p-series, so the series √ converges by the Direct Comparison
n=2
n3/2 n=2 n5 +5
Test.
214 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)


∞
1
48.
n=1
3n − 2n

solution Apply the Limit Comparison Test with an = 1

3n −2n and bn = 1
3n . Then,

an 3n 1
L = lim = lim n = lim  =1
n→∞ 3 − 2n
n
n→∞ bn n→∞
1− 2 3

∞
1
The series n
is a convergent geometric series; because L exists, we may therefore conclude by the Limit
n=1
3
∞
1
Comparison Test that the series also converges.
n=1
3 − 2n
n


∞
n10 + 10n
49.
n=1
n11 + 11n

n10 +10n 10 n
solution Apply the Limit Comparison Test with an = n11 +11n
and bn = 11 . Then,

n10 +10n n10 +10n n10

an +1
= lim n +11 = lim 10n 10n
11 n
L = lim n → ∞ = lim =1
bn n→∞ 10 n n→∞ n11 +11n n→∞ n11 +1
11 11n 11n

∞ 
 n
10
The series is a convergent geometric series; because L exists, we may therefore conclude by the
n=1
11

∞
n10 + 10n
Limit Comparison Test that the series also converges.
n=1
n11 + 11n


∞
n20 + 21n
50.
n=1
n21 + 20n

n20 +21n 21 n
solution Apply the Limit Comparison Theorem with an = n21 +20n
and bn = 20 . Then

n20 +21n n20 +21n n20

an +1
= lim n +20n = lim
21 n 21n 21n
L = lim n21 +20n
= lim =1
n→∞ bn n→∞ 21 n→∞
20 20n
n→∞ n21
20n +1

∞ 
 n
21
The series is a divergent geometric series. Since L = 1, the two series either both converge or both
n=1
20

∞
n20 + 21n
diverge; thus, we may conclude from the Limit Comparison Test that the series diverges.
n=1
n21 + 20n


∞  n
2n + n 2
51. Determine the convergence of using the Limit Comparison Test with bn = .
n=1
3n − 2 3

2n +n
solution With an = 3n −2 , we have
 n
an 2n + n 3n 6n + n3n 1 + n 12
L = lim = lim n · = lim n = lim  =1
n→∞ 3 − 2 2n n→∞ 6 − 2n+1
n
n→∞ bn n→∞
1−2 1 3

∞  n
 2
Since L = 1, the two series either both converge or both diverge. Since is a convergent geometric
n=1
3

∞
2n + n
series, the Limit Comparison Test tells us that also converges.
n=1
3n − 2
 Chapter Review Exercises 215

∞
ln n 1
52. Determine the convergence of n
using the Limit Comparison Test with bn = .
n=1
1.5 1.4n
ln n
solution With an = , and using L’Hôpital’s Rule,
1.5n
ln n
an 1.5n ln n
L = lim = lim = lim 
n→∞ bn n→∞ 1 n→∞ 1.5 n
1.4n 1.4

1.4 n
1/n 1 1.5
= lim  = lim =0
n→∞ 1.5 n ln(1.5/1.4) n→∞ n
ln(1.5/1.4) 1.4


∞
Since L < ∞ and bn is a convergent geometric series, it follows from the Limit Comparison Test that
n=1
∞
ln n
n
also converges.
n=1
1.5

∞
1
53. Let an = 1 − 1 − 1n . Show that lim an = 0 and that an diverges. Hint: Show that an ≥ .
n→∞
n=1
2n
solution
√ √
1 n−1 n− n−1 n − (n − 1) 1
1− 1− =1− = √ = √ √ √ = √
n n n n( n + n − 1) n + n2 − n
1 1
≥ √ =
n+ n2 2n
 #  
∞
1
The series diverges, so the series ∞ n=2 1 − 1 − 1
n also diverges by the Direct Comparison Test.
n=2
2n
⎛ ⎞
∞ ⎜
⎜⎜⎜ 1 ⎟⎟⎟⎟
54. Determine whether ⎜⎝1 − 1 − 2 ⎟⎠ converges.
n=2
n
solution
√ 
1 n2 − 1 n − n2 − 1 n2 − n2 − 1
1− 1− 2 =1− = =  √
n n2 n n n + n2 − 1
1 1 1
=
√  = √ ≤ 2
n n+ n −1 2 n +n n −1 n
2 2

⎛ ⎞
∞
1 ∞ ⎜
⎜⎜⎜ 1 ⎟⎟⎟⎟
The series is a convergent p-series, so the series ⎜⎝1 − 1 − 2 ⎟⎠ also converges by the Direct
n=1
n2 n=2
n
Comparison Test.
∞
n
55. Consider 2 + 1)2
.
n=1
(n
(a) Show that the series converges.
(b) Use the inequality in (4) from Exercise 83 of Section 10.3 with M = 99 to approximate the sum
of the series. What is the maximum size of the error?
solution
(a) For n ≥ 1,
n n 1
< 22 = 3
(n2 + 1) 2 (n ) n
∞
1 ∞
n
The series 3
is a convergent p-series, so the series 2 + 1)2
also converges by the Direct Compari-
n=1
n n=1
(n
son Test.
216 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

(b) With an = n
(n2 +1)2
, f (x) = x
(x2 +1)2
and M = 99, Eq. (4) in Exercise 83 of Section 10.3 becomes


99
n ∞
x  100
n ∞
x
+ dx ≤ S ≤ + dx
n=1
(n2 + 1)2 100 (x2 + 1)2 n=1
(n2 + 1)2
100 (x2 + 1)2
or
⎛ 99 ⎞
⎜⎜⎜ n ∞
x ⎟⎟⎟ 100
0 ≤ S − ⎜⎜⎜⎝ + dx ⎟⎟⎟ ≤
⎠
n=1
(n + 1)2
2
100 (x + 1)
2 2 (1002 + 1)2

Now,

99
n
= 0.397066274; and
n=1
(n2 + 1)2
∞ R  
x x 1 1 1
dx = lim dx = lim − 2 +
100 (x + 1)2
2 R→∞ 100 (x2 + 1)2 2 R→∞ R + 1 1002 + 1
1
= = 0.000049995
20002
thus,

S ≈ 0.397066274 + 0.000049995 = 0.397116269

The bound on the error in this approximation is

100
= 9.998 × 10−7
(1002 + 1)2

In Exercises 56–59, determine whether the series converges absolutely. If it does not, determine whether it
converges conditionally.
∞
(−1)n
56. √3
n=1
n + 2n
√ √
solution Both 3 n and 2n are increasing functions, so 3 n + 2n is also increasing. Therefore, √
3
1
n+2n
is
decreasing. Moreover,
1
lim √3 =0
n→∞ n + 2n
∞
(−1)n
so the series √3 converges by the Alternating Series Test.
n=1
n + 2n
∞
1
The corresponding positive series is √3 . Because
n=1
n + 2n

1 1 1 1
√3 > = ·
n + 2n n + 2n 3 n

∞
1 
∞
1
and the harmonic series diverges, √3 also diverges by the Direct Comparison Test. Thus,
n n + 2n
#∞ (−1)n
n=1 n=1

n=1 √
3
n+2n
converges conditionally.

∞
(−1)n
57.
n=1
n1.1 ln(n + 1)

∞
1
solution Consider the corresponding positive series . Because
n=1
n1.1 ln(n + 1)

1 1
<
n1.1 ln(n + 1) n1.1
 Chapter Review Exercises 217

 ∞
1 ∞
(−1)n
and is a convergent p-series, we can conclude by the Direct Comparison Test that
n=1
n1.1
n=1
n ln(n + 1)
1.1

 ∞
(−1) n
also converges. Thus, 1.1 ln(n + 1)
converges absolutely.
n=1
n

∞ cos π + πn
4
58. √
n=1
n
solution Note
π  √
π π n 2
cos + πn = cos cos nπ − sin sin nπ = (−1)
4 4 4 2
Therefore,

 π  2  (−1)n
∞ cos ∞ ∞
4 + πn (−1)n 2
√ = √ √ = √ √
n=1
n n=1
n 2 2 n=1 n


∞ cos π
4 + πn
Now, the sequence { √1n } is decreasing and converges to 0 as n → ∞. Therefore, √converges
n=1
n
by the Alternating Series Test. However, the corresponding positive series is a divergent p-series (p = 12 ), so
the original series converges conditionally.

∞ cos π + 2πn
4
59. √
n=1
n
 √
solution cos π4 + 2πn = cos π4 = 22 , so
 √ 
∞ cos π + 2πn ∞
4 2 1
√ = √
n=1
n 2 n=1
n

∞ cos π + 2πn
4
This is a divergent p-series, so the series √ diverges.
n=1
n
 ∞
(−1)n
60. Use a computer algebra system to approximate 3+
√ to within an error of at most 10−5 .
n=1
n n
#
solution The sequence { n3 +1√n } is decreasing and converges to 0, so the series ∞ (−1)n
√
n=1 n3 + n converges by the
Alternating Series Test. Using the Error Bound for an alternating series,
 
 ∞
(−1)n  1
S N − √  ≤ aN+1 = √
 n=1
n + n
3
(N + 1) + N + 1
3

If we want an approximation with an error of at most 10−5 , we must choose N such that
1 √
√ < 10−5 or (N + 1)3 +
N + 1 > 105
(N + 1)3 + N + 1
√ √
For N = 45, (N + 1)3 + N + 1 = 97342.8 < 105 , and for N = 46, (N + 1)3 + N + 1 = 103829.9 > 105 .
The smallest acceptable value for N is therefore N = 46. Using a computer algebra system, we find
∞
(−1)n
√ ≈ S 46 = −0.418452236
n=1
n3 + n


∞
(−1)k
61. Catalan’s constant is defined by K = .
k=0
(2k + 1)2
(a) How many terms of the series are needed to calculate K with an error of less than 10−6 ?
(b) Carry out the calculation.
218 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

solution
(a) Using the Error Bound for an alternating series, we have
1 1
|S N − K| ≤ =
(2(N + 1) + 1)2 (2N + 3)2
For accuracy to 10−6 , we must choose N so that
1
< 10−6 or (2N + 3)2 > 106
(2N + 3)2
Solving for N yields
997
N> = 498.5
2
(b) Computing the sum using a CAS gives

499
(−1)k
K≈ = 0.915965094
k=0
(2k + 1)2


∞ 
∞ 
∞
62. Give an example of conditionally convergent series an and bn such that (an + bn ) converges
n=1 n=1 n=1
absolutely.
(−1)n (−1)n+1
solution Let an = n and bn = n .
The corresponding alternating series converge by the Alternating

∞
Series Test; however, the corresponding positive series are the divergent harmonic series. Thus, an and
n=1

∞
bn converge conditionally. On the other hand, the series
n=1

∞ ∞    ∞    ∞
(−1)n (−1)n+1 n 1 −1
(an + bn ) = + = (−1) + = 0
n=1 n=1
n n n=1
n n n=1

converges absolutely.

∞
63. Let an be an absolutely convergent series. Determine whether the following series are convergent or
n=1
divergent:
∞   
∞
1
(a) an + 2 (b) (−1)n an
n=1
n n=1
∞
1 ∞
|an |
(c) (d)
n=1
1 + a2n n=1
n

∞ 
∞ 
∞
solution Because an converges absolutely, we know that an converges and that |an | converges.
n=1 n=1 n=1

∞ ∞
1
(a) Because we know that an converges and the series 2
is a convergent p-series, the sum of these
n
∞   n=1 n=1
 1
two series, an + 2 also converges.
n=1
n
(b) We have,

∞ 
∞
|(−1)n an | = |an |
n=1 n=1


∞ 
∞ 
∞
Because |an | converges, it follows that (−1)n an converges absolutely, which implies that (−1)n an
n=1 n=1 n=1
converges.
 Chapter Review Exercises 219


∞
(c) Because an converges, limn→∞ an = 0. Therefore,
n=1

1 1
lim = =10
n→∞ 1 + a2n 1 + 02

∞
1
and the series diverges by the Divergence Test.
n=1
1 + a2n
∞ ∞
|an |
|an |
(d) n ≤ |an | and the series |an | converges, so the series also converges by the Direct Comparison
n=1 n=1
n
Test.
√
64. Let {an } be a positive sequence such that lim n an = 12 . Determine whether the following series converge
n→∞
or diverge:
∞ ∞  ∞
√
(a) 2an (b) 3n an (c) an
n=1 n=1 n=1
solution
(a)
n √n √ 1 1
L = lim 2 n an = 1 · =
2an = lim
n→∞ n→∞ 2 2
Because L < 1, the series converges by the Root Test.
(b)
n √ 1 3
L = lim 3n an = lim 3 n an = 3 · =
n→∞ n→∞ 2 2
Because L > 1, the series diverges by the Root Test.
(c)
√ √n 1
L = lim an = lim an =
n

n→∞ n→∞ 2
Because L < 1, the series converges by the Root Test.

In Exercises 65–72, apply the Ratio Test to determine convergence or divergence, or state that the Ratio Test
is inconclusive.
∞
n5
65.
n=1
5n
n5
solution With an = 5n ,
   5
 an+1  = (n + 1) · 5 = 1 1 + 1
5 n
 an  5n+1 n5 5 n
and
   5
 an+1  1 1 1 1
ρ = lim   = lim 1 + = ·1=
n→∞ an  5 n→∞ n 5 5
Because ρ < 1, the series converges by the Ratio Test.
∞ √
n+1
66.
n=1
n8
√
solution With an = n+1
n8
,
  √
 an+1  = n + 2 · n
8
n + 2  n 8
 an  (n + 1)8 √ =
n+1 n+1 n+1
and
 
an+1 
ρ = lim   = 1 · 18 = 1
n→∞ an 

Because ρ = 1, the Ratio Test is inconclusive.

220 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)


∞
1
67.
n=1
n2n + n3

solution With an = 1
n2n +n3
,

  2
n2n 1 + n2n
 an+1  =
2
n2n + n3 1 n 1 + n2n
 an  (n + 1)2n+1 + (n + 1)3 =  = · ·
(n + 1)2n+1 1 + (n+1)
2
2 n + 1 1 + (n+1) 2

2n+1 2n+1

and
 
an+1  1
ρ = lim 
1
 = · 1 · 1 =
n→∞ an 2 2
Because ρ < 1, the series converges by the Ratio Test.

∞
n4
68.
n=1
n!
4
solution With an = nn! ,
 
 an+1  = (n + 1) · n! = (n + 1)
4 3
an+1
 an  (n + 1)! n4 and ρ = lim =0
n4 n→∞ an
Because ρ < 1, the series converges by the Ratio Test.

∞
2n
2

69.
n=1
n!
n2
solution With an = 2n! ,
   
 an+1  = 2 an+1 
(n+1)2
22n+1
ρ = lim 
n!
 an  (n + 1)! · 2n2 = n + 1 and =∞
n→∞ an 

Because ρ > 1, the series diverges by the Ratio Test.

∞
ln n
70.
n=4
n3/2

solution With an = ln n
n3/2
,
   n 3/2 ln(n + 1)
 an+1  = ln(n + 1) · n =
3/2
 an  (n + 1)3/2 ln n n+1 ln n
and
 
an+1 
ρ = lim   = 13/2 · 1 = 1
n→∞ an 

Because ρ = 1, the Ratio Test is inconclusive.

∞  n
n 1
71.
n=1
2 n!
 n
solution With an = n2 n1! ,
   n+1  n  n  n
 an+1  = n + 1 1
·
2
n! =
1 n+1
=
1
1+
1
 an  2 (n + 1)! n 2 n 2 n
and
 
an+1  1
ρ = lim  = e
n→∞ an  2
Because ρ = e
2 > 1, the series diverges by the Ratio Test.
 Chapter Review Exercises 221

∞  n
n 1
72.
n=1
4 n!

n n 1
solution With an = 4 n!
,
   n+1  n  n  n
 an+1  = n + 1 1
·
4
n! =
1 n+1
=
1
1 +
1
 an  4 (n + 1)! n 4 n 4 n
and
 
an+1  1
ρ = lim  = e
n→∞ an  4
Because ρ = e
4 < 1, the series converges by the Ratio Test.
In Exercises 73–76, apply the Root Test to determine convergence or divergence, or state that the Root Test
is inconclusive.
∞
1
73. n
n=1
4
solution With an = 1
4n ,

√n n 1 1
L = lim an = lim n
=
n→∞ n→∞ 4 4
Because L < 1, the series converges by the Root Test.
∞  n
2
74.
n=1
n
 n
solution With an = 2n ,
 
n
2 2
L = lim = lim = 0
n

n→∞ n n→∞ n

Because L < 1, the series converges by the Root Test.

∞  n
3
75.
n=1
4n
 n
solution With an = 4n 3
,
 
n
√n 3 3
L = lim an = lim = lim =0
n

n→∞ n→∞ 4n n→∞ 4n

Because L < 1, the series converges by the Root Test.

∞  n3
1
76. cos
n=1
n
 n3
solution With an = cos 1n ,

 n3  n2   x2
√n n 1 1 1
L = lim an = lim cos = lim cos = lim cos
n→∞ n→∞ n n→∞ n x→∞ x
Now,
   
  ln cos 1x
1
− sin 1x − x12
1 cos ( 1
)
ln L = lim x2 ln cos = lim = lim x

x→∞ x x→∞
x
1
2
x→∞ − x23

1 1 sin 1x 1 1
= − lim  · lim =− ·1·1=−
2 x→∞ cos 1 x→∞ 1x 2 2
x

Therefore, L = e−1/2 . Because L < 1, the series converges by the Root Test.
222 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

In Exercises 77–100, determine convergence or divergence using any method covered in the text.
∞  n
2
77.
n=1
3

solution This is a geometric series with ratio r = 2

3 < 1; hence, the series converges.

∞
π7n
78.
n=1
e8n
π7
solution This is a geometric series with ratio r = e8
≈ 1.013, so it diverges.

∞
79. e−0.02n
n=1

solution This is a geometric series with common ratio r = 1

e0.02
≈ 0.98 < 1; hence, the series converges.

∞
80. ne−0.02n
n=1

solution With an = ne−0.02n ,

 
 an+1  = (n + 1)e
−0.02(n+1)
n + 1 −0.02
 an  −0.02n
= e
ne n
and
 
an+1 
ρ = lim   = 1 · e−0.02 = e−0.02
n→∞ an 

Because ρ < 1, the series converges by the Ratio Test.


∞
(−1)n−1
81. √ √
n=1 n+ n+1
solution In this alternating series, an = √ √
1
n+ n+1
. The sequence {an } is decreasing, and

lim an = 0
n→∞

therefore the series converges by the Alternating Series Test.


∞
1
82.
n=10
n(ln n)3/2

solution Let f (x) = x(ln1x)3/2 . This function is continuous, positive and decreasing for x ≥ 10; therefore, the
Integral Test applies. Now,
∞ R ln R
dx dx ln 10
3/2
= lim 3/2
= lim du
10 x(ln x) R→∞ 10 x(ln x) R→∞ 1 u3/2
 ln R ⎞  
−2  ⎟⎟⎟⎟ 1
= ln 10 · lim √  ⎟⎠ = 2 ln 10 lim 1 − √ = 2 ln 10
R→∞ u 1 R→∞ ln R
The integral converges; hence, the series converges as well.

∞
(−1)n
83.
n=2
ln n

solution The sequence an = 1

ln n is decreasing for n ≥ 10 and

lim an = 0
n→∞

therefore, the series converges by the Alternating Series Test.

 Chapter Review Exercises 223

∞
n!
84.
n=1
(2n)!
solution Use the Ratio Test:
   
 an+1   (n + 1)! (2n)!  n+1
lim   = lim  ·  = lim
n→∞  an  n→∞ (2(n + 1))! n!  n→∞ (2n + 2)(2n + 1)
n+1 1 + 1/n
= lim = lim =0
n→∞ 4n2 + 6n + 2 n→∞ 4n + 6 + 2/n
So by the Ratio Test, the series converges.
∞
n
85.
n=2
1 + 100n
solution Note that
n 1 1
lim = lim =
n→∞ 1 + 100n n→∞ 1/n + 100 100
Therefore the terms of the sequence do not converge to zero, so this series is divergent since it fails the
Divergence Test.
∞
n3 − 2n2 + n − 4
86.
n=2
2n4 + 3n3 − 4n2 − 1
n3 −2n2 +n−4
solution Use the Limit Comparison Test with an = 2n4 +3n3 −4n2 −1
and bn = 1n . Note that an is positive for
n ≥ 3, and
an n(n3 − 2n2 + n − 4) n4 − 2n3 + n − 4 1 − 2/n + 1/n2 − 4/n4 1
lim = lim 4 = lim = lim =
n→∞ bn n→∞ 2n + 3n3 − 4n2 − 1 n→∞ 2n4 + 3n3 − 4n2 − 1 n→∞ 2 + 3/n − 4/n2 − 1/n4 2
∞
Since the limit is 12 > 0 and bn diverges, being the Harmonic Series, it follows from the Limit Com-
∞ n=2
parison Test that an diverges as well.
n=2

∞
cos n
87.
n=1
n3/2
solution Note that this is not a positive series, since cos n may be either positive or negative. However,
|cos n| ≤ 1, so that
 
 cos n  ≤ 1
 n3/2  n3/2
∞ 1
But is a p-series with p = 32 > 1, so it is convergent. Thus the original series is absolutely
n=1 n3/2
convergent by the Direct Comparison Test.
∞
n
88. √
n=1 n +1
3/2

solution Note that

n 1
lim √ = lim √ =∞
n→∞ n3/2 + 1 n→∞ n−1/2 + n−2
Therefore the terms of the sequence do not converge to zero, so this series is divergent since it fails the
Divergence Test.
∞ 
 n n
89.
n=1
5n + 2
∞  1 n
solution Since 5n + 2 > 5n, it follows that 5n+2 < 5n = 5 . Since
n n 1
is a geometric series with
n=1 5
1
ratio 5 , it converges, so the original series converges as well by the Direct Comparison Test.
224 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)


∞
en
90.
n=1
n!
n
solution With an = ne! ,
   
 an+1  = e an+1 
n+1
n! e
ρ = lim  =0
 an  (n + 1)! · en = n + 1 and
n→∞ an 

Because ρ < 1, the series converges by the Ratio Test.

∞
1
91. √
n=1 n n + ln n

solution For n ≥ 1,
1 1 1
√ ≤ √ = 3/2
n n + ln n n n n
∞
1 ∞
1
The series 3/2
is a convergent p-series, so the series √ converges by the Direct Comparison
n=1 n n + ln n
n=1
n
Test.
 ∞
1
92. √3 √
n=1
n(1 + n)
solution Apply the Limit Comparison Test with an = √
3
1 √
n(1+ n)
and bn = 1
n5/6
. Then,

√ 1 √
3
n(1+ n) n5/6 1
L= lim = lim √3 = lim =1
n→∞ 1
n5/6
n→∞ n + n5/6 n→∞ √1
n
+1

∞
1 ∞
1
The series is a divergent p-series. Because L > 0, the series √3 √ also diverges by the
n=1
n5/6
n=1
n(1 + n)
Limit Comparison Test.
∞  
1 1
93. √ − √
n=1
n n+1
solution This series telescopes:
 ∞        
1 1 1 1 1 1 1
√ − √ = 1 − √ + √ − √ + √ − √ + ...
n=1
n n+1 2 2 3 3 4

so that the nth partial sum S n is

       
1 1 1 1 1 1 1 1
Sn = 1 − √ + √ − √ + √ − √ + ··· + √ − √ =1− √
2 2 3 3 4 n n+1 n+1
and then
∞  
1 1 1
√ − √ = lim S n = 1 − lim √ =1
n=1
n n+1 n→∞ n→∞ n+1


∞
94. ln n − ln(n + 1)
n=1

solution This series telescopes:


∞
ln n − ln(n + 1) = (ln 1 − ln 2) + (ln 2 − ln 3) + (ln 3 − ln 4) + . . .
n=1

so that the nth partial sum S n is

S n = (ln 1 − ln 2) + (ln 2 − ln 3) + (ln 3 − ln 4) + · · · + (ln n − ln(n + 1))
= ln 1 − ln(n + 1) = − ln(n + 1)
 Chapter Review Exercises 225

and then

∞
ln n − ln(n + 1) = lim S n = − lim ln(n + 1) = ∞
n→∞ n→∞
n=1

so the series diverges.


∞
1
95. √
n=1
n+ n

solution For n ≥ 101, the sequence 1

n−100.1 is positive and

1 1
>
n − 100.1 n
∞
1 ∞
1
Now, diverges, so diverges by the Direct Comparison Test. Since adding the finite
n=101
n n=101
n − 100.1

100
1 ∞
1
sum does not affect the convergence of the series, it follows that diverges as
n=1
n − 100.1 n=1
n − 100.1
well.
 ∞
cos(πn)
96.
n=2
n2/3

solution cos(πn) = (−1)n , so


∞
cos(πn) 
∞
(−1)n
=
n=2
n2/3 n=2
n2/3

The sequence an = 1
n2/3
is decreasing and

lim an = 0
n→∞

therefore, the series converges by the Alternating Series Test.

∞
1
97.
n=2
n n
ln

solution For n ≥ N large enough, ln n ≥ 2 so that

∞
1 ∞
1
≤
n=N
nln n
n=N
n2

∞
1
which is a convergent p-series. Thus by the Comparison Test, also converges; adding back in the
n=N
n n
ln

terms for n < N does not affect convergence.

∞
1
98. 3
n=2 ln n

solution For N large enough, ln n ≤ n1/4 when n ≥ N so that

∞
1 ∞
1
3
> 3/4
n=N ln n n=N
n

∞
1
which is a divergent p-series. Thus by the Comparison Test, 3
diverges; adding back in the terms for
n=N ln n
n < N does not affect this result.
226 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)


∞
π
99. sin2
n=1
n
solution For all x > 0, sin x < x. Therefore, sin2 x < x2 , and for x = πn ,

π π2 1
<sin2 = π2 · 2
n n2 n
∞
1 ∞
π
The series 2
is a convergent p-series, so the series sin2 also converges by the Direct Comparison
n=1
n n=1
n
Test.
∞
22n
100.
n=0
n!
2n
solution With an = 2n! ,
   
 an+1  = 2 an+1 
2(n+1)
n! 4 
 =0
 an  (n + 1)! · 22n = n + 1 and ρ = lim 
n→∞ an 

Because ρ < 1, the series converges by the Ratio Test.

In Exercises 101–106, find the interval of convergence of the power series.

∞
2n xn
101.
n=0
n!
2n xn
solution With an = n! ,
   n+1 n+1 
 an+1   2 x n!  2
ρ = lim   = lim  · n n  = |x| lim =0
n→∞ an  n→∞  (n + 1)! 2 x  n→∞ n + 1

The radius of convergence is therefore R = ∞, and the series converges for all x.
∞
xn
102.
n=0
n+1
xn
solution With an = n+1 ,
   n+1 
 an+1   x n  n
ρ = lim   = lim  ·  = |x| lim = |x|
n→∞ an  n→∞  n + 1 xn  n→∞ n + 1

Thus ρ < 1 when |x| < 1, so the radius of convergence is R = 1, and the series converges absolutely for
∞
1 ∞
1
−1 < x < 1. For the endpoint x = 1, the series becomes = , which is the divergent harmonic
n=0
n + 1 n=1 n
∞
(−1)n
series. For the endpoint x = −1, the series becomes , which converges by the Alternating Series
n=0
n+1
∞
xn
Test. The series therefore converges for −1 ≤ x < 1.
n=0
n+1

∞
n6
103. (x − 3)n
n=0
n8 +1
n6 (x−3)n
solution With an = n8 +1
,
   
 an+1   (n + 1)6 (x − 3)n+1 n8 + 1 
ρ = lim   = lim  · 6 
n→∞ an  n→∞  (n + 1)8 + 1 n (x − 3)n 
(n + 1)6 (n8 + 1) n14 + · · ·
= |x − 3| lim = |x − 3| lim = |x − 3|
n→∞ ((n + 1)8 + 1)n6 n→∞ n14 + · · ·

Thus ρ < 1 when |x − 3| < 1, so the radius of convergence is R = 1 centered at x = 3, and the series converges
∞
n6
absolutely for 2 < x < 4. For the endpoint x = 4, the series becomes , which converges by the
n=0
n +1
8
 Chapter Review Exercises 227

∞
1
Direct Comparison Test comparing with the convergent p-series . For the endpoint x = 2, the series
n=1
n2
∞
n6 (−1)n ∞
n6 (x − 3)n
becomes , which converges by the Alternating Series Test. The series therefore
n=0
n8 + 1 n=0
n8 + 1
converges for 2 ≤ x ≤ 4.

∞
104. nxn
n=0

solution With an = nxn ,

   
 an+1   (n + 1)xn+1  n+1
ρ = lim   = lim   = |x| n→∞
lim = |x|
n→∞ an  n→∞  nxn n

Thus ρ < 1 when |x| < 1, so the radius of convergence is R = 1, and the series converges absolutely for
 ∞
−1 < x < 1. For the endpoint x = 1, the series becomes n, which diverges by the Divergence Test. For
#∞ n=0
the endpoint x = −1, the series becomes n=0 (−1)n n, which also diverges by the Divergence Test. The series
∞
nxn therefore converges for −1 < x < 1.
n=0


∞
105. (nx)n
n=0

solution With an = nn xn ,
      n 
an+1   (n + 1)n+1 xn+1  (n + 1)n+1
ρ = lim 
1
 = lim   = |x| lim = |x| lim (n + 1) 1 + = ∞ if x  0
n→∞ an n→∞  nn xn n→∞ nn n→∞ n

Thus ρ = ∞ unless x = 0, so the series converges only at x = 0.


∞
(2x − 3)n
106.
n=2
n ln n
(2x−3)n
solution With an = n ln n ,

   
 an+1   (2x − 3)n+1 n ln n 
ρ = lim   = lim  · 
n→∞ an  n→∞  (n + 1) ln(n + 1) (2x − 3)n 
 
n ln n n ln n
= |2x − 3| lim = |2x − 3| lim ·
n→∞ (n + 1) ln(n + 1) n→∞ n + 1 ln(n + 1)

By L’Hôpital’s Rule,
ln n 1/n n+1
lim = lim = lim =1
n→∞ ln(n + 1) n→∞ 1/(n + 1) n→∞ n
so that both factors in the above limit approach 1 as n → ∞. Thus ρ = |2x − 3|, so that ρ < 1 when |2x − 3| < 1.
This gives

−1 < 2x − 3 < 1 ⇒ 2 < 2x < 4 ⇒ 1<x<2

Therefore the radius of convergence is centered at x = 32 , and the series converges absolutely for 1 <
1
2
∞
1
x < 2. For the endpoint x = 2, the series becomes , which diverges by the Integral Test. For the
n=2
n ln n
∞
(−1)n
endpoint x = −1, the series becomes , which converges by the Alternating Series Test. The series
n=2
n ln n
∞
(2x − 3)n
therefore converges for 1 ≤ x < 2.
n=2
n ln n
228 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

2
107. Expand f (x) = as a power series centered at c = 0. Determine the values of x for which the
4 − 3x
series converges.
solution Write
2 1 1
=
4 − 3x 2 1 − 34 x

Substituting 34 x for x in the Maclaurin series for 1

1−x ,
we obtain
∞  n
1 3
= xn
1 − 34 x n=0 4
 
This series converges for  34 x < 1, or |x| < 43 . Hence, for |x| < 43 ,
∞  n
2 1 3
= xn
4 − 3x 2 n=0 4

108. Prove that


∞
e−x
ne−nx =
n=0
(1 − e−x )2

Hint: Express the left-hand side as the derivative of a geometric series.


∞ ∞
solution For x > 0, e−nx = (e−x )n is a convergent geometric series with ratio r = e−x ; hence,
n=0 n=0


∞
1
e−nx =
n=0
1 − e−x

Differentiating term-by-term then yields


∞
e−x
−ne−nx = −
n=0
(1 − e−x )2

Therefore, for x > 0,


∞
e−x
ne−nx =
n=0 (1 − e−x )2

∞
x2k
109. Let F(x) = .
k=0
2k · k!
(a) Show that F(x) has infinite radius of convergence.
(b) Show that y = F(x) is a solution of

y = xy + y, y(0) = 1, y (0) = 0

(c) Plot the partial sums S N for N = 1, 3, 5, 7 on the same set of axes.
solution
x2k
(a) With ak = 2k ·k!
,
 
 ak+1  = |x|2k+2 2k · k! x2
 ak  2k+1 · (k + 1)! · |x|2k = 2(k + 1)

and
 
ak+1 
ρ = lim   = x2 · 0 = 0
n→∞ ak 

Because ρ < 1 for all x, we conclude that the series converges for all x; that is, R = ∞.
 Chapter Review Exercises 229

(b) Let
∞
x2k
y = F(x) =
k=0
2k · k!

Then

∞
2kx2k−1 
∞
x2k−1
y = =
k=1
2k k! k=1
2k−1 (k− 1)!

∞
(2k − 1)x2k−2
y =
k=1
2k−1 (k − 1)!

and

∞
x2k−1  x2k  ∞
x2k  x2k ∞ ∞
xy + y = x + = +1+
k=1
2 (k − 1)! k=0 2 k! k=1 2 (k − 1)!
k−1 k k−1
k=1
2k k!

∞
(2k + 1)x2k 
∞
(2k + 1)x2k 
∞
(2k − 1)x2k−2
=1+ = = = y
k=1
2k k! k=0
2k k! k=1
2k−1 (k − 1)!

Moreover,
∞
02k 
∞
02k−1
y(0) = 1 + = 1 and y (0) = =0
k=1
2k k! k=1
2k−1 (k − 1)!
∞
x2k
Thus, k k!
is the solution to the equation y = xy + y satisfying y(0) = 1, y (0) = 0.
k=0
2
(c) The partial sums S 1 , S 3 , S 5 , and S 7 are plotted in the figure below.
y
7
6
5
4
3
2
1
x
−2 −1 1 2


∞
110. Find a power series P(x) = an xn that satisfies the Laguerre differential equation
n=0

xy + (1 − x)y − y = 0
with initial condition satisfying P(0) = 1.
solution Let

∞
y = P(x) = an xn
n=0
Then,

∞ 
∞
y = nan xn−1 , y = n(n − 1)an xn−2
n=1 n=2
and

∞ 
∞ 
∞ 
∞
xy + (1 − x)y − y = n(n − 1)an xn−1 + nan xn−1 − nan xn − an xn
n=2 n=1 n=1 n=0


∞ 
∞ 
∞ 
∞
= (n + 1)nan+1 xn + (n + 1)an+1 xn − nan xn − an xn
n=1 n=0 n=1 n=0


∞  
= (a1 − a0 ) + (n + 1)2 an+1 − (n + 1)an xn
n=1
230 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

In order for this series to be equal to zero, the coefficient of xn must be equal to zero for each n; thus
an
a1 = a0 and an+1 =
n+1
Now, y(0) = P(0) = a0 , so to satisfy the initial condition P(0) = 1, we must set a0 = 1. Then,

a1 = a0 = 1;
a1 1
a2 = = ;
2 2
a2 1 1
a3 = = = ;
3 6 3!
a3 1
a4 = =
4 4!
and, in general, an = n1! . Thus,

∞
xn
P(x) = = ex
n=0
n!

x2 e x
111. Use power series to evaluate lim .
x→ 0 cos x − 1

solution The relevant Maclaurin series are

x2 x3 x4
ex = 1 + x + + + + ···
2! 3! 4!
x2 x4 x6
cos x = 1 − + − + ···
2! 4! 6!
Hence the limit becomes

x2 x3 x4
x2 e x x2 1 + x + 2! + 3! + 4! + ···
lim = lim 
x→0 cos x − 1 x→0 1 − 2!
x 2
+ 4!
x 4
− 6!
x 6
+ ··· − 1

x2 x3 x4
x2 1 + x + 2! + 3! + 4! + ···
= lim
− 2! + 4! − 6! + ···
x 2 x 4 x 6
x→0

x2 x3 x4
1+x+ 2! + 3! + 4! + ···
= lim
x→0 − 2!1 + x2
4! − x4
6! + ···
1
= = −2
−1/2

x2 (1 − ln(x + 1))
112. Use power series to evaluate lim .
x→ 0 sin x − x
solution The relevant Maclaurin series are

x2 x3 x4
ln(x + 1) = x − + − + ···
2 3 4
x3 x5 x7
sin x = x − + − + ···
3! 5! 7!
Hence the limit becomes
  2 3 4
x2 (1 − ln(x + 1)) x2 1 − x − x2 + x3 − x4 + · · ·
lim = lim 
x→0 sin x − x x→0 x − 3!
x3
+ 5!
x5
− 7!
x7
+ ··· − x
  2 3 4
x2 1 − x − x2 + x3 − x4 + · · ·
= lim
x→0 − 3!
x3
+ 5!
x5
− 7!
x7
+ ···
 Chapter Review Exercises 231

x2 x3 x4
1−x+ 2 − 3 + 4 − ···
= lim
x→0 − 3!x + x3
5! − x5
7! + ···
2 3 4
1 − x + x2 − x3 + x4 − · · ·
= lim 
x→0 − 3!x − 5!
x3
+ 7!
x5
− ···

If x → 0+ , the denominator is always negative, so the right-hand limit is −∞. If x → 0− , the denominator is
always positive, so the left-hand limit is ∞. The limit does not exist.

In Exercises 113–118, find the Taylor polynomial at x = a for the given function.

113. f (x) = x3 , T3, a=1

solution We start by computing the first three derivatives of f (x) = x3 :

f  (x) = 3x2
f  (x) = 6x
f  (x) = 6

Evaluating the function and its derivatives at x = 1, we find

f (1) = 1, f  (1) = 3, f  (1) = 6, f  (1) = 6

Therefore,
f  (1) f  (1)
T 3 (x) = f (1) + f  (1)(x − 1) + (x − 2)2 + (x − 1)3
2! 3!
6 6
= 1 + 3(x − 1) + (x − 2)2 + (x − 1)3
2! 3!
= 1 + 3(x − 1) + 3(x − 2)2 + (x − 1)3

114. f (x) = 3(x + 2)3 − 5(x + 2), T3, a = −2

solution T 3 (x) is the Taylor polynomial of f consisting of powers of (x + 2) up to three. Since f (x) is
already in this form we conclude that T 3 (x) = f (x).
115. f (x) = x ln(x), T4, a=1
solution We start by computing the first four derivatives of f (x) = x ln x:
1
f  (x) = ln x + x · = ln x + 1
x
1
f  (x) =
x
1
f  (x) = −
x2
2
f (4) (x) = 3
x
Evaluating the function and its derivatives at x = 1, we find

f (1) = 0, f  (1) = 1, f  (1) = 1, f  (1) = −1, f (4) (1) = 2

Therefore,
f  (1) f  (1) f (4) (1)
T 4 (x) = f (1) + f  (1)(x − 1) + (x − 1)2 + (x − 1)3 + (x − 1)4
2! 3! 4!
1 1 2
= 0 + 1(x − 1) + (x − 1)2 − (x − 1)3 + (x − 1)4
2! 3! 4!
1 1 1
= (x − 1) + (x − 1)2 − (x − 1)3 + (x − 1)4
2 6 12
232 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

116. f (x) = (3x + 2)1/3 , T3, a = 2

solution We start by computing the first three derivatives of f (x) = (3x + 2)1/3 :
1
f  (x) = (3x + 2)−2/3 · 3 = (3x + 2)−2/3
3
2
f  (x) = − (3x + 2)−5/3 · 3 = −2(3x + 2)−5/3
3
10
f  (x) = (3x + 2)−8/3 · 3 = 10(3x + 2)−8/3
3
Evaluating the function and its derivatives at x = 2, we find
1  1 5
f (2) = 2, f  (2) = , f (2) = − , f  (2) =
4 16 128
Therefore,
f  (2) f  (2)
T 3 (x) = f (2) + f  (2)(x − 2) + (x − 2)2 + (x − 2)3
2! 3!
1 −1/16 5/128
= 2 + (x − 2) + (x − 2)2 + (x − 2)3
4 2! 3!
1 1 5
= 2 + (x − 2) − (x − 2)2 − (x − 2)3
4 32 768

117. f (x) = xe−x ,

2
T4, a=0
We start by computing the first four derivatives of f (x) = xe−x :
2
solution

f  (x) = e−x + x · (−2x)e−x = (1 − 2x2 )e−x

2 2 2

f  (x) = −4xe−x + (1 − 2x2 ) · (−2x)e−x = (4x3 − 6x)e−x

2 2 2

f  (x) = (12x2 − 6)e−x + (4x3 − 6x) · (−2x)e−x = (−8x4 + 24x2 − 6)e−x
2 2 2

f (4) (x) = (−32x3 + 48x)e−x + (−8x4 + 24x2 − 6) · (−2x)e−x = (16x5 − 80x3 + 60x)e−x
2 2 2

Evaluating the function and its derivatives at x = 0, we find

f (0) = 0, f  (0) = 1, f  (0) = 0, f  (0) = −6, f (4) (0) = 0
Therefore,
f  (0) 2 f  (0) 3 f (4) (0) 4
T 4 (x) = f (0) + f  (0)x + x + x + x
2! 3! 4!
6
= 0 + x + 0 · x2 − x3 + 0 · x4 = x − x3
3!

118. f (x) = ln(cos x), T 3 , a=0

solution We start by computing the first three derivatives of f (x) = ln(cos x):
sin x
f  (x) = − = − tan x
cos x
f  (x) = −sec2 x
f  (x) = −2 sec2 x tan x
Evaluating the function and its derivatives at x = 0, we find
f (0) = 0, f  (0) = 0, f  (0) = −1, f  (0) = 0
Therefore,
f  (0) 2 f  (0) 3 0 1 0 x2
T 3 (x) = f (0) + f  (0)x + x + x = 0 + x − x2 + x3 = −
2! 3! 1! 2! 3! 2
 Chapter Review Exercises 233

119. Find the nth Maclaurin polynomial for f (x) = e3x .

solution We differentiate the function f (x) = e3x repeatedly, looking for a pattern:

f  (x) = 3e3x = 31 e3x

f  (x) = 3 · 3e3x = 32 e3x
f  (x) = 3 · 32 e3x = 33 e3x

Thus, for general n, f (n) (x) = 3n e3x and f (n) (0) = 3n . Substituting into the formula for the nth Taylor polyno-
mial, we obtain:
3x 32 x2 33 x3 34 x4 3n xn
T n (x) = 1 + + + + + ··· +
1! 2! 3! 4! n!
1 1 1
= 1 + 3x + (3x)2 + (3x)3 + · · · + (3x)n
2! 3! n!
√
120. Use the fifth Maclaurin polynomial of f (x) = e x to approximate e. Use a calculator to determine the
error.
solution Let f (x) = e x . Then f (n) (x) = e x and f (n) (0) = 1 for all n. Hence,

f  (0) 2 f  (0) 3 f (4) (0) 4 f (5) (0) 5

T 5 (x) = f (0) + f  (0)x + x + x + x + x
2! 3! 4! 5!
x2 x3 x4 x5
=1+x+ + + +
2! 3! 4! 5!
For x = 1
2 we have

   2  3  4  5
1 1 1 1
1 1 2 2 2 2
T5 =1+ + + + +
2 2 2! 3! 4! 5!
1 1 1 1 1
=1+ + + + + = 1.648697917
2 8 48 384 3840
√
Using a calculator, we find that e = 1.648721271. The error in the Taylor polynomial approximation is

|1.648697917 − 1.648721271| = 2.335 × 10−5

121. Use the third Taylor polynomial of f (x) = tan−1 x at a = 1 to approximate f (1.1). Use a calculator to
determine the error.
solution We start by computing the first three derivatives of f (x) = tan−1 x:

1
f  (x) =
1 + x2
2x
f  (x) = −
1 + x2 2
 2  
−2 1 + x2 + 2x · 2 1 + x2 · 2x 2 3x2 − 1
f  (x) = =
1 + x2 4 1 + x2 3
Evaluating the function and its derivatives at x = 1, we find
π  1 1 1
f (1) = , f (1) = , f  (1) = − , f  (1) =
4 2 2 2
Therefore,
f  (1) f  (1)
T 3 (x) = f (1) + f  (1)(x − 1) + (x − 1)2 + (x − 1)3
2! 3!
π 1 1 1
= + (x − 1) − (x − 1)2 + (x − 1)3
4 2 4 12
234 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

Setting x = 1.1 yields:

π 1 1 1
T 3 (1.1) = + (0.1) − (0.1)2 + (0.1)3 = 0.832981496
4 2 4 12
Using a calculator, we find tan−1 1.1 = 0.832981266. The error in the Taylor polynomial approximation is
 
T (1.1) − tan−1 1.1 = |0.832981496 − 0.832981266| = 2.301 × 10−7
3

√
122. Let T 4 be the Taylor polynomial for f (x) = x at a = 16. Use the Error Bound to find the maximum
possible size of | f (17) − T 4 (17)|.
solution Using the Error Bound, we have

(17 − 16)5 K
| f (17) − T 4 (17)| ≤ K =
5! 5!
  √
where K is a number such that  f (5) (x) ≤ K for all 16 ≤ x ≤ 17. Starting from f (x) = x we find
1 −1/2  1 3 15
f  (x) = x , f (x) = − x−3/2 , f  (x) = x−5/2 , f (4) (x) = − x−7/2
2 4 8 16
and
105 −9/2
f (5) (x) = x
32
For 16 ≤ x ≤ 17,
 (5)  105 105 105
 f (x) = ≤ =
32x9/2 32 · 169/2 8388608
Therefore, we may take
105
K=
8388608
Finally,
105 1
| f (17) − T 4 (17)| ≤ · ≈ 1.044 · 10−7
8388608 5!

123. Find n such that |e − T n (1)| < 10−8 , where T n is the nth Maclaurin polynomial for f (x) = e x .
solution Using the Error Bound, we have

|1 − 0|n+1 K
|e − T n (1)| ≤ K =
(n + 1)! (n + 1)!
 
where K is a number such that  f (n+1) (x) = e x ≤ K for all 0 ≤ x ≤ 1. Since e x is increasing, the maximum
value on the interval 0 ≤ x ≤ 1 is attained at the endpoint x = 1. Thus, for 0 ≤ u ≤ 1, eu ≤ e1 < 2.8. Hence
we may take K = 2.8 to obtain:
2.8
|e − T n (1)| ≤
(n + 1)!
We now choose n such that
2.8
< 10−8
(n + 1)!
(n + 1)!
> 108
2.8
(n + 1)! > 2.8 × 108
For n = 10, (n + 1)! = 3.99 × 107 < 2.8 × 108 and for n = 11, (n + 1)! = 4.79 × 108 > 2.8 × 108 . Hence, to
make the error less than 10−8 , n = 11 is sufficient; that is,
|e − T 11 (1)| < 10−8
 Chapter Review Exercises 235

124. Let T 4 be the Taylor polynomial for f (x) = x ln x at a = 1 computed in Exercise 115. Use the Error
Bound to find a bound for | f (1.2) − T 4 (1.2)|.
solution Using the Error Bound, we have

(1.2 − 1)5 (0.2)5

| f (1.2) − T 4 (1.2)| ≤ K = K
5! 120
 
where K is a number such that  f (5) x ≤ K for all 1 ≤ x ≤ 1.2. Starting from f (x) = x ln x, we find

1 1 1 2
f  (x) = ln x + x = ln x + 1, f  (x) = , f  (x) = − 2 , f (4) (x) = 3
x x x x
and
−6
f (5) (x) =
x4
For 1 ≤ x ≤ 1.2,
 (5)  6 6
 f (x) = 4 ≤ 4 = 6
x 1
Hence we may take K = 6 to obtain:

(0.2)5
| f (1.2) − T 4 (1.2)| ≤ 6 = 1.6 × 10−5
120

125. Verify that T n (x) = 1 + x + x2 + · · · + xn is the nth Maclaurin polynomial of f (x) = 1/(1 − x). Show
using substitution that the nth Maclaurin polynomial for f (x) = 1/(1 − x/4) is

1 1 1
T n (x) = 1 + x + 2 x2 + · · · + n xn
4 4 4
1
What is the nth Maclaurin polynomial for g(x) = ?
1+x
solution Let f (x) = (1 − x)−1 . Then, f  (x) = (1 − x)−2 , f  (x) = 2(1 − x)−3 , f  (x) = 3!(1 − x)−4 , and, in
general, f (n) (x) = n!(1 − x)−(n+1) . Therefore, f (n) (0) = n! and

1! 2! n!
T n (x) = 1 + x + x2 + · · · + xn = 1 + x + x2 + · · · + xn
1! 2! n!
1
Upon substituting x/4 for x, we find that the nth Maclaurin polynomial for f (x) = is
1 − x/4

1 1 1
T n (x) = 1 + x + 2 x2 + · · · + n xn
4 4 4
1
Substituting −x for x, the nth Maclaurin polynomial for g(x) = is
1+x

T n (x) = 1 − x + x2 − x3 + − · · · + (−x)n

5
126. Let f (x) = and let ak be the coefficient of xk in the Maclaurin polynomial T n for k ≤ n.
4 + 3x − x2
 
1/4 1
(a) Show that f (x) = + .
1 − x/4 1 + x
1
(b) Use Exercise 125 to show that ak = + (−1)k .
4k+1
(c) Compute T 3 .
236 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

solution
(a) Start by factoring the denominator and writing the form of the partial fraction decomposition:
5 5 A B
f (x) = = = +
4 + 3x − x2 (x + 1)(4 − x) x + 1 4 − x
Multiplying through by (x + 1)(4 − x), we obtain:

5 = A(4 − x) + B(x + 1)

Substituting x = 4 yields 5 = A(0) + B(5), so B = 1; substituting x = −1 yields 5 = A(5) + B(0), so A = 1.

Thus,
1
1 1 1
f (x) = + = + 4
x+1 4−x x+1 1− x
4
1
(b) The nth Maclaurin polynomial for f (x) = 4
1− 4x + 1
x+1 is the sum of the nth Maclaurin polynomials for the
functions g(x) = · 1
4
1
and h(x) =
1− 4x
1
1+x .
In Exercise 125, we found that the nth Maclaurin polynomials Pn (x)
and Qn (x) for g and h are
   n
1 1 1 2 1 n 1 1 1 1 xk
Pn (x) = 1 + x + 2 x + · · · + n x = + 2 x + 3 x2 + · · · + n+1 xn =
4 4 4 4 4 4 4 4 k=0
4k+1

n
Qn (x) = 1 − x + x2 − x3 + · · · + (−1)n xn = (−1)k xk
k=0

Therefore,
n 
n n % &
xk 1
T n (x) = Pn (x) + Qn (x) = + (−1) x =
k k
+ (−1) xk
k

k=0
4k+1 k=0 k=0
4k+1

that is, the coefficient of xk in T n for k ≤ n is

1
ak = + (−1)k
4k+1

solution
(a) Let f (x) = sin x + sinh x. Then

f  (x) = cos x + cosh x

f  (x) = − sin x + sinh x
f  (x) = − cos x + cosh x
f (4) (x) = sin x + sinh x

From this point onward, the pattern of derivatives repeats indefinitely. Thus

f (0) = f (4) (0) = f (8) (0) = sin 0 + sinh 0 = 0

f  (0) = f (5) (0) = cos 0 + cosh 0 = 2
f  (0) = f (6) (0) = − sin 0 + sinh 0 = 0
f  (0) = f (7) (0) = − cos 0 + cosh 0 = 0

Consequently,

f (5) (0) 5 1 5
T 5 (x) = f  (0)x + x = 2x + x
5! 60
and, because f (6) (0) = f (7) (0) = f (8) (0) = 0, it follows that
1 5
T 6 (x) = T 7 (x) = T 8 (x) = T 5 (x) = 2x + x
60
 Chapter Review Exercises 237

(b) First note that | sin x| ≤ 1 and | cos x| ≤ 1 for all x. Moreover,
 x 
e − e−x  e x + e−x
| sinh x| =  ≤ = cosh x
2  2
Now, recall from part (a), that all derivatives of f (x) contain two terms: the first is ± sin x or ± cos x, while
the second is either sinh x or cosh x. In absolute value, the trigonometric term is always less than or equal to
1, while the hyperbolic term is always less than or equal to cosh x. Thus, for all n,
f (n) (x) ≤ 1 + cosh x
(c) Using the Error Bound, we have
K|x − 0|9 K|x|9
|T 8 (x) − f (x)| ≤ =
9! 9!
 
where K is a number such that  f (u) ≤ K for all u between 0 and x. By part (b), we know that
(9)

f (9) (u) ≤ 1 + cosh u

Now, cosh u is an even function that is increasing on (0, ∞). The maximum value for u between 0 and x
is therefore cosh x. Moreover, for −1 ≤ x ≤ 1, cosh x ≤ cosh 1 ≈ 1.543 < 1.6. Hence, we may take
K = 1 + 1.6 = 2.6, and
2.6 9
|T 8 (x) − f (x)| ≤ |x|
9!

In Exercises 127–136, find the Taylor series centered at c.

127. f (x) = e4x , c=0
solution Substituting 4x for x in the Maclaurin series for e x yields

∞
(4x)n 
∞
4n
e4x = = xn
n=0
n! n=0
n!

128. f (x) = e2x , c = −1

solution Write:

e2x = e2(x+1)−2 = e−2 e2(x+1)

Substituting 2(x + 1) for x in the Maclaurin series for e x yields

∞
(2(x + 1))n 
∞
2n
e2(x+1) = = (x + 1)n
n=0
n! n=0
n!

hence,

∞
2n (x + 1)n
e2x = e−2
n=0
n!

129. f (x) = x4 , c=2

solution We have

f  (x) = 4x3 f  (x) = 12x2 f  (x) = 24x f (4) (x) = 24

and all higher derivatives are zero, so that
f (2) = 24 = 16 f  (2) = 4 · 23 = 32 f  (2) = 12 · 22 = 48 f  (2) = 24 · 2 = 48 f (4) (2) = 24
Thus the Taylor series centered at c = 2 is
4
f (n) (2) 32 48 48 24
(x − 2)n = 16 + (x − 2) + (x − 2)2 + (x − 2)3 + (x − 2)4
n=0
n! 1! 2! 3! 4!

= 16 + 32(x − 2) + 24(x − 2)2 + 8(x − 2)3 + (x − 2)4

238 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

130. f (x) = x3 − x, c = −2
solution We have

f  (x) = 3x2 − 1 f  (x) = 6x f  (x) = 6

and all higher derivatives are zero, so that

f (−2) = −8 + 2 = −6 f  (−2) = 3(−2)2 − 1 = 11 f  (−2) = 6(−2) = −12 f  (−2) = 6

Thus the Taylor series centered at c = −2 is

3
f (n) (−2) 11 −12 6
(x + 2)n = −6 + (x + 2) + (x + 2)2 + (x + 2)3
n=0
n! 1! 2! 3!

= −6 + 11(x + 2) − 6(x + 2)2 + (x + 2)3

131. f (x) = sin x, c=π

solution Because sin x = − sin(x − π),

x sin x = (x − π) sin x + π sin x = −(x − π) sin(x − π) − π sin(x − π)

Substituting x − π for x in the Maclaurin for sin x yields


∞
(x − π)2n+1
sin(x − π) = (−1)n
n=0
(2n + 1)!

Thus,

∞
(x − π)2n+1 ∞
(x − π)2n+1
x sin x = −(x − π) (−1)n −π (−1)n
n=0
(2n + 1)! n=0
(2n + 1)!

∞
(x − π)2n+2 ∞
(x − π)2n+1
= (−1)n+1 +π (−1)n+1
n=0
(2n + 1)! n=0
(2n + 1)!

132. f (x) = e x−1 , c = −1

solution Write

e x−1 = e x+1−1−1 = e−2 e x+1

Substituting x + 1 for x in the Maclaurin series for e x yields


∞
(x + 1)n
e x+1 =
n=0
n!

hence,

∞
(x + 1)n 
∞
(x + 1)n
e x−1 = e−2 =
n=0
n! n=0
n!e2

1
133. f (x) = , c = −2
1 − 2x
solution Write
1 1 1 1
= =
1 − 2x 5 − 2(x + 2) 5 1 − 25 (x + 2)

Substituting 25 (x + 2) for x in the Maclaurin series for 1

1−x yields

1  2n ∞
= (x + 2)n
1 − 5 (x + 2) n=0 5n
2
 Chapter Review Exercises 239

hence,
1  2n  2n
∞ ∞
1
= (x + 2)n = (x + 2)n
1 − 2x 5 n=0 5n
n=0
5n+1

1
134. f (x) = , c = −2
(1 − 2x)2
solution Note that
d 1 2
=
dx 1 − 2x 1 − 2x
1
so that we can derive the Taylor series for f (x) by differentiating the Taylor series for 1−2x , computed in the
previous exercise, and dividing by 2. Thus
⎛∞ ⎞
1 1 d ⎜⎜⎜⎜ 2n n⎟
⎟
= · ⎜⎝ (x + 2) ⎟⎟⎟⎠
(1 − 2x) 2 2 dx n=0 5 n+1

1  n2n 2  n2n−1
∞ ∞
= (x + 2) n−1
= (x + 2)n−1
2 n=1 5n+1 50 n=1 5n−1

1  (k + 1)2k
∞
= (x + 2)k
25 k=0 5k

x
135. f (x) = ln , c=2
2
solution Write
   
x (x − 2) + 2 x−2
ln = ln = ln 1 +
2 2 2

Substituting x−2
2 for x in the Maclaurin series for ln(1 + x) yields

n+1 x−2 n
x  (−1) 
∞ ∞
2 (−1)n+1 (x − 2)n
ln = =
2 n=1 n n=1
n · 2n

This series is valid for |x − 2| < 2.

 x
136. f (x) = x ln 1 + , c = 0
2
solution Substituting 2x for x in the Maclaurin series for ln(1 + x) yields
 n
 x   (−1)
∞ n−1 x ∞
(−1)n−1 xn
2
ln 1 + = =
2 n=1
n n=1
n2n

Thus,
 x ∞
(−1)n−1 xn  (−1)n−1 xn+1
∞
x ln 1 + =x =
2 n=1
n2n n=1
n2n

In Exercises 137–140, find the first three terms of the Maclaurin series of f (x) and use it to calculate f (3) (0).
2
137. f (x) = (x2 − x)e x
solution Substitute x2 for x in the Maclaurin series for e x to get
2 1 4 1 6
e x = 1 + x2 + x + x + ...
2 6
so that the Maclaurin series for f (x) is
2 1 6 1
(x2 − x)e x = x2 + x4 + x + · · · − x − x3 − x5 − · · · = −x + x2 − x3 + x4 + . . .
2 2
240 C H A P T E R 10 INFINITE SERIES (LT CHAPTER 11)

The coefficient of x3 is
f  (0)
= −1
3!
so that f  (0) = −6.
138. f (x) = tan−1 (x2 − x)
solution Substitute x2 − x for x in the Maclaurin series for tan−1 x to get
1 1
tan−1 (x2 − x) = (x2 − x) − (x2 − x)3 + · · · = −x + x2 + x3 + . . .
3 3
The coefficient of x3 is
f  (0) 1
=
3! 3
so that f  (0) = 3! 31 = 2.
1
139. f (x) =
1 + tan x
solution Substitute − tan x in the Maclaurin series for 1
1−x to get
1
= 1 − tan x + (tan x)2 − (tan x)3 + . . .
1 + tan x
We have not yet encountered the Maclaurin series for tan x. We need only the terms up through x3 , so compute
tan (x) = sec2 x tan (x) = 2(tan x) sec2 x tan (x) = 2(1 + tan2 x) sec2 x + 4(tan2 x) sec2 x
so that

tan (0) = 1 tan (0) = 0 tan (0) = 2

Then the Maclaurin series for tan x is

tan (0) tan (0) 2 tan (0) 3 1
tan x = tan 0 + x+ x + x + · · · = x + x3 + . . .
1! 2! 3! 3
Substitute these into the series above to get
   2  3
1 1 1 1
= 1 − x + x3 + x + x3 − x + x3 + . . .
1 + tan x 3 3 3
1 3
=1−x− x + x2 − x3 + higher degree terms
3
4
= 1 − x + x2 − x3 + higher degree terms
3
The coefficient of x3 is
f  (0) 4
=−
3! 3
so that
4
f  (0) = −6 · = −8
3
√
140. f (x) = (sin x) 1 + x
√
solution The binomial series for 1 + x is

      
√ 1/2 1/2 1/2 2 1/2 3
1 + x = (1 + x) = 1/2
+ x+ x + x + ...
0 1 2 3
  
2 −2 − 12 − 32
1 1 1
1 2
=1+ x+ x2 + x3 + . . .
2 2 3!
1 1 1 3
= 1 + x − x2 + x + ...
2 8 16
 Chapter Review Exercises 241

So, multiply the first few terms of the two Maclaurin series together:
  
√ x3 1 1 1 3
(sin x) 1 + x = x − 1 + x − x2 + x
6 2 8 16
1 2 1 3 1 3
= x+ x − x − x + higher-degree terms
2 8 6
1 7 3
= x + x2 − x + higher-degree terms
2 24
The coefficient of x3 is
f  (0) 7
=−
3! 24
so that
7 7
f  (0) = −6 · =−
24 4

π π3 π5 π7
141. Calculate − 3 + 5 − 7 + ··· .
2 2 3! 2 5! 2 7!
solution We recognize that

π π3 π5 π7  (π/2)2n+1
∞
− 3 + 5 − 7 + ··· = (−1)n
2 2 3! 2 5! 2 7! n=0
(2n + 1)!

is the Maclaurin series for sin x with x replaced by π/2. Therefore,

π π3 π5 π7 π
− 3 + 5 − 7 + · · · = sin = 1
2 2 3! 2 5! 2 7! 2
x
et − 1
142. Find the Maclaurin series of the function F(x) = dt.
0 t
solution Subtracting 1 from the Maclaurin series for et yields
∞ n
t ∞ n
t ∞ n
t
et − 1 = −1=1+ −1=
n=0
n! n=1
n! n=1
n!

Thus,

et − 1 1  tn  tn−1
∞ ∞
= =
t t n=1 n! n=1 n!

Finally, integrating term-by-term yields

x
∞ n−1 
∞ ∞
x
et − 1 t x n−1
t xn
dt = dt = dt =
0 t 0 n=1
n! n=1 0 n! n=1
n! n