Section 7.

2
C07S02.001: You don’t need l’Hôpital’s rule to evaluate this limit, but you may use it:

x−1 1 1
lim = lim = .
x→1 x2 − 1 x→1 2x 2

C04S08.002: You don’t need l’Hôpital’s rule to evaluate this limit, but you may use it:

3x − 4 3 3
lim = lim = .
x→∞ 2x − 5 x→∞ 2 2

C07S02.003: You don’t need l’Hôpital’s rule to evaluate this limit, but you may use it (twice):

2x2 − 1 4x 4 2
lim 2
= lim = lim = .
x→∞ 5x + 3x x→∞ 10x + 3 x→∞ 10 5

C07S02.004: You don’t need l’Hôpital’s rule to evaluate this limit (apply the definition of the derivative
to the evaluation of f  (0) where f (x) = e3x ), but you may use it:

e3x − 1 3e3x
lim = lim = 3.
x→0 x x→0 1

C07S02.005: Without l’Hôpital’s rule:

sin x2 sin x2
lim = lim x · = 0 · 1 = 0.
x→0 x x→0 x2
sin x
(We used Theorem 1 of Section 2.3, lim = 1, and the product law for limits.)
x→0 x
With l’Hôpital’s rule:

sin x2 2x cos x2
lim = lim = 2 · 0 · 1 = 0.
x→0 x x→0 1

C07S02.006: You don’t need l’Hôpital’s rule to evaluate this limit (see the solution to Problem 3 of Section
2.3), but you may use it:
√ 1 −1/2
1 − cos x 2x sin x1/2 sin x1/2 1
lim = lim+ = lim+ 1/2
=
x→0 + x x→0 1 x→0 2x 2
by Theorem 1 of Section 2.3. If you prefer a “pure” l’Hôpital’s rule solution, you should substitute x = u2
to obtain
√
1 − cos x 1 − cos u sin u cos u 1
lim = lim 2
= lim = lim = .
x→0 + x u→0 u u→0 2u u→0 2 2
Note that it was necessary to apply l’Hôpital’s rule twice in the second solution.

C07S02.007: You may not use l’Hôpital’s rule! The numerator is approaching zero but the denominator
is not. Hence use the quotient law for limits (Section 2.2):

x−1 lim (x − 1) 0
lim = x→1 = = 0.
x→1 sin x lim sin x sin 1
x→1

1
Note that illegal use of l’Hôpital’s rule in this problem will result in the incorrect value sec 1 ≈ 1.8508157177
for the limit.

C07S02.008: The numerator and denominator of the fraction are both approaching zero, so you may try
using l’Hôpital’s rule (twice):

1 − cos x sin x cos x

lim 3
= lim 2
= lim .
x→0 x x→0 3x x→0 6x

But the latter limit does not exist (the left-hand limit is −∞ and the right-hand limit is +∞). So the
hypotheses of Theorem 1 of Section 7.2 are not satisfied; this is a case in which l’Hôpital’s rule (as stated in
Theorem 1) has failed. Other measures are needed. By l’Hôpital’s rule (or by Problem 3 of Section 2.3), we
have

1 − cos x sin x cos x 1

lim = lim = lim = .
x→0 x2 x→0 2x x→0 2 2
Therefore

1 − cos x 1
3
≈
x 2x
1 − cos x
if x is close to zero. Consequently lim does not exist.
x→0 x3

C04S08.009: Without l’Hôpital’s rule we might need to resort to the Taylor series methods of Section 11.9
to evaluate this limit. But l’Hôpital’s rule may be applied (twice):

ex − x − 1 ex − 1 ex 1
lim = lim = lim = .
x→0 x2 x→0 2x x→0 2 2

C07S02.010: You may not apply l’Hôpital’s rule: The numerator is approaching zero but the denominator
is not. But this means that the quotient law of limits (Section 2.2) may be applied instead:

lim (1 + cos 2z)

1 + cos 2z z→π/2 1−1
lim = = = 0.
z→π/2 1 − sin 2z lim (1 − sin 2z) 1−0
z→π/2

This problem would be more interesting if the denominator were 1 − sin z.

C07S02.011: The numerator and denominator are both approaching zero, so l’Hôpital’s rule may be
applied:

u tan u tan u + u sec2 u

lim = lim
u→0 1 − cos u u→0 sin u
 u 
= lim sec u + · sec2 u = 1 + 1 · 12 = 2.
u→0 sin u
Note that we used the sum law for limits (Section 2.2) and the fact that

u
lim = 1,
u→0 sin u
a consequence of Theorem 1 of Section 2.3 and the quotient law for limits.

2
C07S02.012: Numerator and denominator are both approaching zero, so l’Hôpital’s rule may be applied
(twice):

x − tan x 1 − sec2 x
lim 3
= lim
x→0 x x→0 3x2
   
2 sec2 x tan x 1 sec2 x sin x 1 12 1
= lim − = − lim · = − · ·1 = − .
x→0 6x 3 x→0 cos x x 3 1 3

ln x 1 10
C07S02.013: lim = lim 1 −9/10 = lim = 0.
x1/10 x→∞ x · x1/10
10 x
x→∞ x→∞

C04S08.014: Several applications of l’Hôpital’s rule yield

er er er er er
lim = lim = lim = lim = = +∞.
r→∞ (r + 1)4 r→∞ 4(r + 1)3 r→∞ 12(r + 1)2 r→∞ 24(r + 1) 24
Even though the limit does not exist, the hypotheses of Theorem 1 of Section 7.2 are all satisfied, so the
answer is correct.

ln(x − 9) 1
C07S02.015: lim = lim = 1.
x→10 x − 10 x→10 1 · (x − 9)

t2 + 1 2t 2
C07S02.016: lim = lim = lim −1 = lim (2t) = +∞.
t→∞ t ln t t→∞ 1 + ln t t→∞ t t→∞

C07S02.017: Always verify that the hypotheses of l’Hôpital’s rule are satisfied.

ex + e−x − 2 ex − e−x ex + e−x 1+1

lim = lim = lim = = 1.
x→0 x sin x x→0 x cos x + sin x x→0 2 cos x − x sin x 2−0

C07S02.018: As x → (π/2)− , tan x → +∞ and cos x → 0+ , so that ln(cos x) → −∞. Hence l’Hôpital’s
rule may be tried:

tan x sec2 x cos x − sec x

lim = lim = lim = −∞
x→(π/2)− ln(cos x) x→(π/2)− − sin x x→(π/2)− sin x

because sin x → 1 and sec x → +∞ as x → (π/2)− .

C07S02.019: Methods of Section 2.3 may be used, or l’Hôpital’s rule yields

sin 3x 3 cos 3x 3·1 3

lim = lim = = .
x→0 tan 5x x→0 5 sec2 5x 5·1 5

ex − e−x ex + e−x
C07S02.020: lim = lim = 2.
x→0 x x→0 1
C07S02.021: The factoring techniques of Section 2.2 work well here, or l’Hôpital’s rule yields

x3 − 1 3x2 3
lim = lim = .
x→1 x − 1 x→1 2x
2 2

x3 − 8 3x2 12 3
C07S02.022: lim = lim = = .
x→2 x4 − 16 x→2 4x3 32 8

3
C07S02.023: Both numerator and denominator approach +∞ as x does, so we may attempt to find the
limit with l’Hôpital’s rule:

x + sin x 1 + cos x
lim = lim ,
x→∞ 3x + cos x x→∞ 3 − sin x
but the latter limit does not exist (and so the equals mark in the previous equation is invalid). The reason:
As x → +∞, x runs infinitely many times through numbers of the form nπ where n is a positive even
integer. At such real numbers the value of

1 + cos x
f (x) =
3 − sin x
is 23 . But x also runs infinitely often through numbers of the form nπ where n is a positive odd integer. At
these real numbers the value of f (x) is 0. Because f (x) takes on these two distinct values infinitely often as
x → +∞, f (x) has no limit as x → +∞.
This does not imply that the limit given in Problem 22 does not exist. (Read Theorem 1 carefully.)
In fact, the limit does exist, and we have here the rare phenomenon of failure of l’Hôpital’s rule. Other
techniques must be used to solve this problem. Perhaps the simplest is this:

sin x
x + sin x 1+ 1+0 1
lim = lim x
x→∞ 3x + cos x cos x = 3 − 0 = 3 .
x→∞
3−
x

C07S02.024: First we try l’Hôpital’s rule:

(x2 + 4)1/2 1
2 (x
2
+ 4)−1/2 · 2x x
lim = lim = lim
x→∞ x x→∞ 1 x→∞ (x + 4)1/2
2

1 (x2 + 4)1/2
= lim = lim .
x→∞ 1 (x2
2 + 4)−1/2 · 2x x→∞ x
Another failure of l’Hôpital’s rule! Here are two ways to find the limit.
 2  2
(x2 + 4)1/2 (x2 + 4)1/2 x2 + 4 2x 1
lim = lim = lim 2
= lim = lim = 1.
x→∞ x x→∞ x x→∞ x x→∞ 2x x→∞ 1
Therefore the original limit is also 1. Second method:
 1/2  1/2
(x2 + 4)1/2 x2 + 4 4 √
lim = lim = lim 1+ 2 = 1 = 1.
x→∞ x x→∞ x2 x→∞ x

Methods of Section 11.9 may also be used, but—while quite general—are a bit unwieldy in this case.

2x − 1 2x ln 2 ln 2
C07S02.025: lim = lim = ≈ 0.6309297536.
x→0 3 −1
x x→0 x
3 ln 3 ln 3

C07S02.026: You may apply l’Hôpital’s rule, but watch what happens:

2x 2x ln 2 2x (ln 2)2 2x (ln 2)3

lim = lim = lim = lim = ··· .
x→∞ 3x x→∞ 3x ln 3 x→∞ 3x (ln 3)2 x→∞ 3x (ln 3)3

If we knew that the original limit existed and was finite, we could conclude that it must be zero, but we
don’t even know that it exists. But

4
  3x  x
3 27
= > 3x > ex ,
2 8
so
 x  x
3 2x 2
lim = +∞, and therefore lim x = lim = 0.
x→∞ 2 x→∞ 3 x→∞ 3

C07S02.027: You can solve this problem without l’Hôpital’s rule, but if you intend to use it you should
probably proceed as follows:

√  2 1/2  1/2
x2 − 1 x −1 x2 − 1
lim √ = lim = lim
x→∞ 4x2 − x x→∞ 4x2 − x x→∞ 4x2 − x

 1/2  1/2  1/2

2x 2 1 1
= lim = lim = = .
x→∞ 8x − 1 x→∞ 8 4 2

C07S02.028: As in the previous solution,

√  1/2  1/2
x3 + x x3 + x 3x2 + 1
lim √ = lim = lim
x→∞ 2x3 − 4 x→∞ 2x3 − 4 x→∞ 6x2
 1/2  1/2
6x 1 1
= lim = =√ .
x→∞ 12x 2 2

ln(1 + x) 1
C07S02.029: lim = lim = 1.
x→0 x x→0 1 · (1 + x)

C07S02.030: It would be easier to establish first that

ln(ln x)
lim = 0,
x→∞ ln x
and it would follow immediately that the given limit is zero as well. But let’s see how well a direct approach
succeeds.

1
ln(ln x) x ln x = lim 1
lim = lim =0
x→∞ x ln x x→∞ 1 + ln x x→∞ (1 + ln x)(x ln x)

because

1 1
< if x > 1.8554
(1 + ln x)(x ln x) x
and 1/x → 0 as x → ∞.

C07S02.031: Three applications of l’Hôpital’s rule yield

2ex − x2 − 2x − 2 2ex − 2x − 2 2ex − 2 2ex 1

lim = lim = lim = lim = .
x→0 x3 x→0 3x2 x→0 6x x→0 6 3

C07S02.032: Three applications of l’Hôpital’s rule yield

5
 sin x − tan x cos x − sec2 x − sin x − 2 sec2 x tan x
lim = lim = lim
x→0 x3 x→0 3x2 x→0 6x
− cos x − 4 sec2 x tan2 x − 2 sec4 x −1 − 0 − 2 1
= lim = =− .
x→0 6 6 2

2 − ex − e−x e−x − ex −e−x − ex 1

C07S02.033: lim 2
= lim = lim =− .
x→0 2x x→0 4x x→0 4 2

e3x − e−3x 3e3x + 3e−3x

C07S02.034: lim = lim = 3.
x→0 2x x→0 2

2x − π 2 1
C07S02.035: lim = lim 2
= = 1.
x→π/2 tan 2x x→π/2 2 sec 2x sec2 π

C07S02.036: The “direct approach” yields

sec x sec x tan x tan x sec2 x sec x

lim = lim 2
= lim = lim = lim = ··· .
x→π/2 tan x x→π/2 sec x x→π/2 sec x x→π/2 sec x tan x x→π/2 tan x

Proceed instead as follows (without l’Hôpital’s rule):

sec x 1 cos x 1 1
lim = lim · = lim = = 1.
x→π/2 tan x x→π/2 cos x sin x x→π/2 sin x 1

x − 2 cos πx 1 + 2π sin πx 1+0 1

C07S02.037: lim = lim = = .
x→2 x2 − 4 x→2 2x 4 4

2x − sin πx 2 − π cos πx 2−π·0 1

C07S02.038: lim = lim = = .
x→1/2 4x2 − 1 x→1/2 8x 4 2

C07S02.039: We first simplify (using laws of logarithms), then apply l’Hôpital’s rule:
 
1
1/2 1
ln(2x) 2 (ln 2 + ln x) 3 x  3
=  lim
1 
lim = lim+ 1 = .
x→0+ ln(3x)1/3 x→0
3 (ln 3 + ln x) 2 x→0+ 2
x

C07S02.040: One application of l’Hôpital’s rule yields

?
ln(1 + x) x−1
lim = lim .
x→0 ln(1 − x2 ) x→0 2x

The limit on the right-hand side does not exist, so we use left-hand and right-hand limits:

ln(1 + x) x−1 ln(1 + x) x−1

lim = lim+ = −∞ and lim = lim− = +∞.
x→0+ ln(1 − x2 ) x→0 2x x→0− ln(1 − x2 ) x→0 2x
Therefore the original limit does not exist.

C07S02.041: Two applications of l’Hôpital’s rule yield

6
 exp(x3 ) − 1 3x2 exp(x3 ) (6x + 9x4 ) exp(x3 )
lim = lim = lim
x→0 x − sin x x→0 1 − cos x x→0 sin x
(6 + 9x3 ) exp(x3 ) (6 + 0) · 1
= lim = = 6.
x→0 sin x 1
x
Alternatively, three applications of l’Hôpital’s rule yield

exp(x3 ) − 1 3x2 exp(x3 ) (6x + 9x4 ) exp(x3 )

lim = lim = lim
x→0 x − sin x x→0 1 − cos x x→0 sin x
(6 + 54x3 + 27x6 ) exp(x3 ) (6 + 54 · 0 + 27 · 0) · 1
= lim = = 6.
x→0 cos x 1

In this problem the Taylor series methods of Section 11.9 are considerably simpler.

C07S02.042: The technique of multiplying numerator and denominator by the conjugate of the numerator
succeeds just as it did in Sections 2.2 and 2.3. Use of l’Hôpital’s rule yields

(1 + 3x)1/2 − 1 1
2 (1 + 3x)−1/2 · 3 3 3
lim = lim = lim = .
x→0 x x→0 1 x→0 2(1 + 3x)1/2 2

(1 + 4x)1/3 − 1 1
3 (1 + 4x)−2/3 · 4 4 4
C07S02.043: lim = lim = lim 2/3
= .
x→0 x x→0 1 x→0 3(1 + 4x) 3

C07S02.044: Multiplication of numerator and denominator by the conjugate of the numerator is one way
to find this limit; l’Hôpital’s rule yields

(3 + 2x)1/2 − (3 + x)1/2 + 2x)−1/2 · 2 − 12 (3 + x)−1/2

1
2 (3
lim = lim
x→0 x x→0 1

1 1 1 1 1
= lim − =√ − √ = √ .
x→0 (3 + 2x)1/2 2(3 + x)1/2 3 2 3 2 3

C07S02.045: If you want to use the conjugate technique to find this limit, you need to know that the
conjugate of a1/3 − b1/3 is a2/3 + a1/3 b1/3 + b2/3 , and the algebra becomes rather long. Here l’Hôpital’s rule
is probably the easy way:

(1 + x)1/3 − (1 − x)1/3 + x)−2/3 + 13 (1 − x)−2/3

1
3 (1
lim = lim
x→0 x x→0 1

1 1 1 1 2
= lim + = + = .
x→0 3(1 + x)2/3 3(1 − x) 2/3 3 3 3

√ 2
1 − tan x − sec2 x − 2 1
C07S02.046: lim = lim = =− .
x→π/4 4x − π x→π/4 4 4 2

ln(1 + x2 ) 2x 0
C07S02.047: lim = lim = = 0.
x→0 ex − cos x x→0 (1 + x2 )(ex + sin x) (1 + 0)(1 + 0)

7
C07S02.048: Because the numerator and denominator are both approaching zero as x → 2, it should be
possible to factor x − 2 out of each, cancel, and proceed without l’Hôpital’s rule. But if we use l’Hôpital’s
rule, the result is

x5 − 5x2 − 12 5x4 − 10x 80 − 20 60 1

lim = lim = = = .
x→2 x10 − 500x − 24 x→2 10x9 − 500 5120 − 500 4620 77

The factor-and-cancel technique yields

x5 − 5x2 − 12 (x − 2)(x4 + 2x3 + 4x2 + 3x + 6)

lim = lim
x→2 x − 500x − 24 x→2 (x − 2)(x + 2x + 4x7 + 8x6 + 16x5 + 32x4 + 64x3 + 128x2 + 256x + 12)
10 9 8

x4 + 2x3 + 4x2 + 3x + 6
= lim
x→2 x9 + 2x8 + 4x7 + 8x6 + 16x5 + 32x4 + 64x3 + 128x2 + 256x + 12
16 + 16 + 16 + 6 + 6 60 1
= = = .
512 + 512 + 512 + 512 + 512 + 512 + 512 + 512 + 512 + 12 4620 77
In this problem l’Hôpital’s rule seems the better choice.

sin2 x
C07S02.049: If f (x) = , then
x
2 sin x cos x
lim f (x) = lim = 2 · 0 · 1 = 0.
x→0 x→0 1
The graph of y = f (x) is next.

0.6

0.4

0.2

-6 -4 -2 2 4 6
-0.2

-0.4

-0.6

C07S02.050: We don’t need l’Hôpital’s rule—we could use Theorem 1 in Section 2.3—but we’ll use the
rule anyway:
   
sin2 x 2 sin x cos x sin x cos x 
lim = lim = lim lim cos x = lim · 1 = 1 · 1 = 1.
x→0 x2 x→0 2x x→0 x x→0 x→0 1

The graph is next.

1

0.8

0.6

0.4

0.2

-6 -4 -2 2 4 6

8
C07S02.051: Here we have

sin x cos x
lim = lim = cos π = −1.
x→π x−π x→π 1
The graph is next.
0.2

-5 -2.5 2.5 5 7.5 10 12.5

-0.2

-0.4

-0.6

-0.8

-1

C07S02.052: One application of l’Hôpital’s rule yields

cos x − sin x 1
lim = lim =− .
x→π/2 2x − π x→π/2 2 2

The graph is next.

0.1

-5 -2.5 2.5 5 7.5

-0.1

-0.2

-0.3

-0.4

-0.5

C07S02.053: Two applications of l’Hôpital’s rule yield

1 − cos x sin x cos x 1

lim 2
= lim = lim = .
x→0 x x→0 2x x→0 2 2
The graph is next.
0.5

0.4

0.3

0.2

0.1

-10 -5 5 10

C07S02.054: Three applications of l’Hôpital’s rule yield

x − sin x 1 − cos x sin x cos x 1

lim 3
= lim 2
= lim = lim = .
x→0 x x→0 3x x→0 6x x→0 6 6

9
The graph is next.

0.15

0.125

0.1

0.075

0.05

0.025

-15 -10 -5 5 10 15

C07S02.055: As x → −∞, f (x) = xe−x → −∞, but

x 1
lim f (x) = lim x
= lim x = 0.
x→∞ x→∞ e x→∞ e

 −x
Also f (x) = (1 − x)e and f (x) = (x − 2)e−x . It follows that the graph of f is increasing for x < 1,


decreasing for x > 1, concave downward for x < 2, and concave upward for x > 2. The positive x-axis is a
horizontal asymptote and the only intercept is (0, 0). The graph of y = f (x) is shown next.

0.2

-1 1 2 3 4 5

-0.2

-0.4

C07S02.056: Given f (x) = x1/2 e−x , we first use l’Hôpital’s rule:

x1/2 1
lim f (x) = lim = lim = 0.
x→∞ x→∞ ex x→∞ 2x1/2 ex

Next,

1 − 2x 4x2 − 4x − 1
f  (x) = and f  (x) = .
2x1/2 ex 4x3/2 ex
So the graph of f√isincreasing for 0 < x < 12 and decreasing for x > 12 . There is an inflection point
where x = 12 1 + 2 ; the y-coordinate is approximately 0.3285738758. The positive x-axis is a horizontal
asymptote and the only intercept is (0, 0). The graph of y = f (x) is shown next.

0.4

0.3

0.2

0.1

1 2 3 4

10
  
C07S02.057: If f (x) = x exp −x1/2 , then

x 2x1/2 2x1/2 2
lim f (x) = lim   = lim   = lim   = lim   = 0.
x→∞ x→∞ exp x 1/2 x→∞ exp x1/2 x→∞ x1/2 exp x1/2 x→∞ exp x1/2

Thus the positive x-axis is a horizontal asymptote. Next,

2 − x1/2 x1/2 − 3
f  (x) =   and f  (x) =  .
2 exp x1/2 4x1/2 exp x1/2

Hence the graph of f is increasing for 0 < x < 4 and decreasing for x > 4; it is concave downward if 0 < x < 9
and concave upward if x > 9. The only intercept is (0, 0). The graph of y = f (x) is shown next.
0.6

0.5

0.4

0.3

0.2

0.1

2.5 5 7.5 10 12.5 15 17.5 20

C07S02.058: Given: f (x) = x2 e−2x . By l’Hôpital’s rule,

x2 2x 2
lim f (x) = lim 2x
= lim 2x
= lim = 0.
x→∞ x→∞ e x→∞ 2e x→∞ 4e2x
So the positive x-axis is a horizontal asymptote. Also note that f (x) → +∞ as x → −∞. Next,

f  (x) = 2x(1 − x)e−2x and f  (x) = 2(2x2 − 4x + 1)e−2x ,

and it follows that the graph

√ of
 f is decreasing if x < 0√and
 if x > 1, increasing if 0 < x < 1; it is concave
upward for x < a = 12 2 − 2 and for x > b = 12 2 + 2 . It is concave downward for a < x < b, so there
are inflection points where x = a and where x = b. The graph of y = f (x) is next.

0.3

0.25

0.2

0.15

0.1

0.05

-1 1 2 3 4

ln x
C07S02.059: Given: f (x) = . So
x
ln x
lim f (x) = lim = −∞
x→0+ +
x→0 x
because the numerator is approaching −∞ and the denominator is approaching 0 through positive values.
Next, using l’Hôpital’s rule,

11
 ln x 1
lim f (x) = lim = lim = 0,
x→∞ x→∞ x x→∞ 1·x
so the positive y-axis is a horizontal asymptote and the negative y-axis is a vertical asymptote. Moreover,

1 − ln x −3 + 2 ln x
f  (x) = and f  (x) = ,
x2 x3

and thus the graph of f is increasing if 0 < x < e, decreasing if x > e, concave downward for 0 < x < e3/2 ,
and concave upward if x > e3/2 . The inflection point where x = e3/2 is not visible because the curvature of
the graph is very small for x > 3. The graph of y = f (x) is next.

0.4

0.2

2 4 6 8
-0.2

-0.4

-0.6

-0.8

-1

C07S02.060: Given:

ln x
f (x) = .
x1/2 + x1/3

We used Mathematica to find f  (x) and solve f  (x) = 0, and thus discovered that there is a global maximum
near (10.094566, 0.433088). Similarly, we found an inflection point near (20.379823, 0.416035). Clearly
f (x) → −∞ as x → 0+ , so the negative y-axis is a vertical asymptote. Also

ln x 1 1
lim f (x) = lim = lim   = lim = 0,
x1/2 + x1/3 x→∞ x 1 x−1/2 + 1 x−2/3 1 1/2
+ 13 x1/3
2x
x→∞ x→∞ x→∞
2 3

so the positive x-axis is a horizontal asymptote. The graph of y = f (x) is next.

0.4

0.3

0.2

0.1

10 20 30 40
-0.1

C07S02.061: The computation in the solution of Problem 55 establishes that

xn
lim =0 (1)
x→∞ ex
in the case n = 1. Suppose that Eq. (1) holds for n = k, a positive integer. Then (by l’Hôpital’s rule)

12
  
xk+1 (k + 1)xk xk
lim = lim = (k + 1) lim = (k + 1) · 0 = 0.
x→∞ ex x→∞ ex x→∞ ex

Therefore, by induction, Eq. (1) holds for every positive integer n. Now suppose that k > 0 (a positive
number not necessarily an integer). Let n be an integer larger than k. Then, for x large positive, we have

xk xn
0 < x
< x.
e e
xk
Therefore, by the squeeze law for limits, lim = 0 for every positive number k.
x→∞ ex
C07s02.062: Suppose that k is a positive real number. Then (by l’Hôpital’s rule)

ln x 1 1
lim = lim = lim = 0.
x→∞ xk x→∞ kxk−1 ·x x→∞ kxk

C07S02.063: Given: f (x) = xn e−x where n is a positive integer larger than 1. Then

xn
lim f (x) = lim =0
x→∞ x→∞ ex
by the result in Problem 61. So the positive x-axis is a horizontal asymptote. Next,

(n − x)xn−1 (x2 − 2nx + n2 − n)xn−2

f  (x) = and f  (x) = .
ex ex
Therefore f  (x) = 0 at the two points (0, 0) and (n, nn e−n ). We consider only the part of the graph for
which x > 0, and the graph of f is increasing for 0 < x < n and decreasing if x > n, so there is a local
√ √
maximum at x = n. Next, f  (x) = 0 when x = a = n − n and when x = b = n + n. It is easy to
establish that f  (x) > 0 if 0 < x < a and if x > b, but that f  (x) < 0 if a < x < b. (Use the fact that
the graph of g(x) = x2 − 2nx + n2 − n is a parabola opening upward.) Therefore the graph of f has two
inflection points for x > 0.

C07S02.064: Given: f (x) = x−k ln x where k is a positive constant. Then

1 − k ln x
f  (x) = ,
xk+1
and the sign of f  (x) is the same as the sign of 1 − k ln x, which is positive if 0 < x < e1/k but negative if
x > e1/k . Hence the graph of f will have a single local maximum where x = e1/k . Next,

k 2 ln x + k ln x − 2k − 1
f  (x) = ,
xk+2
so f  (x) = 0 when
 
2k + 1
x = exp ,
k2 + k

so the graph of f has at most one inflection point. Moreover, if x is near zero then f  (x) < 0, whereas
f  (x) > 0 if x is large positive. Therefore the graph of f has exactly one inflection point. Finally, the result
in Problem 62 shows that the positive x-axis is a horizontal asymptote.

1
C07S02.065: The substitution y = yields
x

13
  
− ln y 1
k
lim x ln x = lim = − lim = 0.
x→0 + y→∞ yk y→∞ ky k

C07S02.066: First suppose that n = −k < 0 where k is a positive integer. Then

(ln x)n 1
lim = lim = 0.
x→∞ x x→∞ x(ln x)k

Next suppose that n = 0. Then

(ln x)n 1
lim = lim = 0.
x→∞ x x→∞ x

Moreover, if n = 1, then by l’Hôpital’s rule,

(ln x)n ln x 1
lim = lim = lim = 0.
x→∞ x x→∞ x x→∞ x

Assume that

(ln x)k
lim = 0
x→∞ x
for some positive integer k. Then, by l’Hôpital’s rule,

(ln x)k+1 (k + 1)(ln x)k (ln x)k

lim = lim = (k + 1) lim = (k + 1) · 0 = 0.
x→∞ x x→∞ x x→∞ x
(ln x)n
Therefore, by induction for positive n, lim = 0 for every integer n.
x→∞ x
C07S02.067: In the following computations we take derivatives with respect to h in the first step. By
l’Hôpital’s rule,

f (x + h) − f (x − h) f  (x + h) + f  (x − h) 2f  (x)
lim = lim = = f  (x).
h→0 2h h→0 2 2
The continuity of f  (x) is needed for two reasons: It implies that f is also continuous, so the first numerator
approaches zero as h → 0; moreover, continuity of f  (x) is needed to ensure that f  (x + h) and f  (x − h)
both approach f  (x) as h → 0.

C07S02.068: In the following computations we take derivatives with respect to h in the first two steps.
By l’Hôpital’s rule,

f (x + h) − 2f (x) + f (x − h) f  (x + h) − f  (x − h)
lim 2
= lim
h→0 h h→0 2h
f  (x + h) + f  (x − h) 2f  (x)
= lim = = f  (x).
h→0 2 2
The continuity of f (x) is needed for the following reasons: We needed to know that f  (x + h) and f  (x − h)


both approach f  (x) as h → 0. We also needed to know that f  was continuous so that the second numerator
approaches zero as h → 0. There is a third reason, which you will see when you discover the reason for the
presence of the term −2f (x) in the first numerator.

C07S02.069: If

14
 (2x − x4 )1/2 − x1/3
f (x) = ,
1 − x4/3

then both the numerator n(x) = (2x − x4 )1/2 − x1/3 and the denominator d(x) = 1 − x4/3 approach zero as
x → 1, and both are differentiable, so l’Hôpital’s rule may be applied. After simplifications we find that

3x2/3 − 6x11/3 − (2x − x4 )1/2 4x1/3

n (x) = and d (x) = − .
3x2/3 (2x − x4 )1/2 3

Therefore

n (x) −3x2/3 + 6x11/3 + (2x − x4 )1/2 −3 + 6 + 1 4

lim f (x) = lim 
= lim 2 = = = 1.
x→1 x→1 d (x) x→1 4x(2x − x )
4 1 4·1 4

C07S02.070: We are to show that if f (x) and g(x) both approach zero as x → +∞, both f  (x) and g  (x)
exist for arbitrarily large values of x, and the second limit in the next line exists, then

f (x) f  (x)
lim = lim  .
x→∞ g(x) x→∞ g (x)

Following the Suggestion, we let F (t) = f (1/t) and G(t) = g(1/t). Then, with t = 1/x, we have

f (x) F (t) F  (t)

lim = lim+ = lim+ 
x→∞ g(x) t→0 G(t) t→0 G (t)

provided that the last limit exists. Note that F and G are differentiable if t > 0 and t is close to zero. Hence

f (x) F  (t) f  (x)

lim = lim+  = lim  .
x→∞ g(x) t→0 G (t) x→∞ g (x)

 x x  x
e2
C07S02.071: lim
lim = lim ex = +∞.
x→∞ e x→∞ e x→∞

C07S02.072: The graph of C(t) for the case A = 1, k = 1, and x = 2 is next.

0.25

0.2

0.15

0.1

0.05

2 4 6 8 10
-0.05

-0.1

Next,
   
dC A x2 1 x2
= √ − exp − .
dt kπt 4kt2 2t 4kt

15
Now dC/dt = 0 when 2x2 t = 4kt2 , so t = 0 or t = x2 /(2k). The general shape of the graph shown here makes
it clear that the former yields the minimum of C(t) (define C(0) = 0 and C will be continuous on [0, +∞])
and the latter yields the maximum, and the maximum pollutant concentration is the corresponding value of
C(t); that is,

A 2
Cmax = .
x πe

C07S02.073: If f (x) = xn e−x (with n a fixed positive integer), then f  (x) = (n − x)xn−1 e−x . Because
f (x)
0 for x
0, f (0) = 0, and f (x) → 0 as x → +∞, f (x) must have a maximum value, and the critical
point where x = n is the sole candidate. Hence the global maximum value of f (x) is f (n) = nn e−n .
Next, f (n − 1) = (n − 1)n e−(n−1) < nn e−n , so
 n
n−1 en−1 1
< = .
n en e

Therefore
 n  −n  −n
n n−1 1
e < = = 1− .
n−1 n n

nn
Also f (n + 1) = (n + 1)n e−(n+1) < . Therefore, by similar computations,
en
 n
1
1+ < e.
n

When we substitute n = 106 (using a computer algebra program, of course) we find that

2.7182804690 < e < 2.7182831877

(round down on the left, up on the right). Thus, to five places, e = 2.71828.

C07S02.074: We used Mathematica 3.0 to plot the graphs of y = ln x and y = x1/10 on the interval [1, 10].
The result is shown next.

1.5

0.5

2 4 6 8 10

The graph makes it clear that a solution of ln x = x1/10 is close to x0 = 3.1. With this initial estimate, a
few iterations of Newton’s method yields the approximate solution x1 ≈ 3.05972667962080885461.

16
 Next we let f (x) = (ln x) − x1/10 and followed the suggestion in Problem 74. The graph of f finally
crossed the x-axis when viewed on the interval [1015 , 1016 ]. A few magnifications yielded the graph shown
next.

0.008

0.006

0.004

0.002

-0.002

-0.004

-0.006

The scale on the x-axis ranges from 3.42×1015 to 3.44×1015 . Thus we have x2 ≈ 3.43×1015 . A few iterations
of Newton’s method soon yielded the more accurate approximation x2 ≈ 3.43063112140780120278 × 1015 .

C07S02.075: Let f (x) = x and g(x) = x2 . The area of the region bounded by the graphs of f and g is
 1
1 1 1
A = (x − x2 ) dx = − = .
0 2 3 6

The moments with respect to the coordinate axes are

 1
1 1 1
My = (x2 − x3 ) dx = − = and
0 3 4 12
 1  
1 2 1 1 1 1
Mx = (x − x4 ) dx = · − = .
0 2 2 3 5 15
Therefore the centroid is located at the point
 
1 2
(x, y) = C , .
2 5

The axis L of rotation is the line y = x; the line through the centroid perpendicular to L has equation

9
−x y =
10
 
9 9
and this perpendicular meets L at the point P , . The distance from P to C is
20 20
 2  2 √
1 9 2 9 2
d = − + − = .
2 20 5 20 20

Because d is the radius of the circle through which C is rotated, the volume generated is (by the first
theorem of Pappus)
√ √
2 1 π 2
V = 2πdA = 2π · · = ≈ 0.07404804897.
20 6 60

17
C07S02.076: We let f (x) = xm and g(x) = xn . Then we used Mathematica 3.0:

A = Integrate[ f[x] − g[x], { x, 0, 1 } ]

1 1
−
1+m 1+n

Then we compute the moments:

My = Integrate[ x∗(f[x] − g[x]), { x, 0, 1 } ];

Mx = Integrate[ (1/2)∗( (f[x])∧2 − (g[x])∧2 ), { x, 0, 1 } ];

Thus the centroid has coordinates

{ xc, yx } = { My/A, Mx/A } // Simplify

(1 + m)(1 + n) 1 + m + n + mn
{ , }
(2 + m)(2 + n) 1 + 2m + 2n + 4mn

For selected values of m and with n = m + 1 we check to see if it’s True that the centroid lies within the
region:

m = 1; n = m + 1;
yc < xc∧m
yc > xc∧n

True
True

m = 2; n = m + 1;
yc < xc∧m
yc > xc∧n

True
True

m = 3; n = m + 1;
yc < xc∧m
yc > xc∧n

False
True

Therefore if m = 3 and n = 4, then the centroid does not lie within the region.

18