# This page intentionally left blank

Student’s Solutions Manual
to accompany Jon Rogawski’s
Single Variable
CALCULUS
SECOND EDITION
BRIAN BRADIE
Christopher Newport University
ROGER LIPSETT
W. H. FREEMAN AND COMPANY
NEW YORK
© 2012 by W. H. Freeman and Company
ISBN-13: 978-1-4292-4290-5
ISBN-10: 1-4292-4290-6
All rights reserved
Printed in the United States of America
First Printing
W. H. Freeman and Company, 41 Madison Avenue, New York, NY 10010
Houndmills, Basingstoke RG21 6XS, England
www.whfreeman.com
CONTENTS
Chapter 1 PRECALCULUS REVIEW 1
1.1 Real Numbers, Functions, and Graphs 1
1.2 Linear and Quadratic Functions 8
1.3 The Basic Classes of Functions 13
1.4 Trigonometric Functions 16
1.5 Technology: Calculators and Computers 23
Chapter Review Exercises 27
Chapter 2 LIMITS 31
2.1 Limits, Rates of Change, and Tangent Lines 31
2.2 Limits: A Numerical and Graphical Approach 37
2.3 Basic Limit Laws 46
2.4 Limits and Continuity 49
2.5 Evaluating Limits Algebraically 57
2.6 Trigonometric Limits 61
2.7 Limits at Inﬁnity 66
2.8 Intermediate Value Theorem 73
2.9 The Formal Deﬁnition of a Limit 76
Chapter Review Exercises 82
Chapter 3 DIFFERENTIATION 91
3.1 Definition of the Derivative 91
3.2 The Derivative as a Function 101
3.3 Product and Quotient Rules 112
3.4 Rates of Change 119
3.5 Higher Derivatives 126
3.6 Trigonometric Functions 132
3.7 The Chain Rule 138
3.8 Implicit Differentiation 147
3.9 Related Rates 157
Chapter Review Exercises 165
Chapter 4 APPLICATIONS OF THE DERIVATIVE 174
4.1 Linear Approximation and Applications 174
4.2 Extreme Values 181
4.3 The Mean Value Theorem and Monotonicity 191
4.4 The Shape of a Graph 198
4.5 Graph Sketching and Asymptotes 206
4.6 Applied Optimization 220
4.7 Newton’s Method 236
4.8 Antiderivatives 242
Chapter Review Exercises 250
Chapter 5 THE INTEGRAL 260
5.1 Approximating and Computing Area 260
5.2 The Deﬁnite Integral 274
5.3 The Fundamental Theorem of Calculus, Part I 284
5.4 The Fundamental Theorem of Calculus, Part II 290
5.5 Net Change as the Integral of a Rate 296
5.6 Substitution Method 300
Chapter Review Exercises 307
Chapter 6 APPLICATIONS OF THE INTEGRAL 317
6.1 Area Between Two Curves 317
6.2 Setting Up Integrals: Volume, Density, Average Value 328
6.3 Volumes of Revolution 336
6.4 The Method of Cylindrical Shells 346
6.5 Work and Energy 355
Chapter Review Exercises 362
Chapter 7 EXPONENTIAL FUNCTIONS 370
7.1 Derivative of f (x) = b
x
and the Number e 370
7.2 Inverse Functions 378
7.3 Logarithms and Their Derivatives 383
7.4 Exponential Growth and Decay 393
7.5 Compound Interest and Present Value 398
7.6 Models Involving y

= k ( y −b) 401
7.7 L’Hôpital’s Rule 407
7.8 Inverse Trigonometric Functions 415
7.9 Hyperbolic Functions 424
Chapter Review Exercises 431
Chapter 8 TECHNIQUES OF INTEGRATION 446
8.1 Integration by Parts 446
8.2 Trigonometric Integrals 457
8.3 Trigonometric Substitution 467
8.4 Integrals Involving Hyperbolic and Inverse Hyperbolic
Functions 481
8.5 The Method of Partial Fractions 485
8.6 Improper Integrals 503
8.7 Probability and Integration 520
8.8 Numerical Integration 525
Chapter Review Exercises 537
Chapter 9 FURTHER APPLICATIONS OF THE
INTEGRAL AND TAYLOR
POLYNOMIALS 555
9.1 Arc Length and Surface Area 555
9.2 Fluid Pressure and Force 564
9.3 Center of Mass 569
9.4 Taylor Polynomials 577
Chapter Review Exercises 593
iii
iv C A L C U L U S CONTENTS
Chapter 10 INTRODUCTION TO DIFFERENTIAL
EQUATIONS 601
10.1 Solving Differential Equations 601
10.2 Graphical and Numerical Methods 614
10.3 The Logistic Equation 621
10.4 First-Order Linear Equations 626
Chapter Review Exercises 637
Chapter 11 INFINITE SERIES 646
11.1 Sequences 646
11.2 Summing an Inﬁnite Series 658
11.3 Convergence of Series with Positive Terms 669
11.4 Absolute and Conditional Convergence 683
11.5 The Ratio and Root Tests 690
11.6 Power Series 697
11.7 Taylor Series 710
Chapter Review Exercises 727
Chapter 12 PARAMETRIC EQUATIONS, POLAR
COORDINATES, AND CONIC
SECTIONS 742
12.1 Parametric Equations 742
12.2 Arc Length and Speed 759
12.3 Polar Coordinates 766
12.4 Area and Arc Length in Polar Coordinates 780
12.5 Conic Sections 789
Chapter Review Exercises 801
June 7, 2011 LTSV SSM Second Pass
1 PRECALCULUS REVIEW
1.1 Real Numbers, Functions, and Graphs
Preliminary Questions
1. Giveanexampleof numbersa andb suchthat a < b and|a| > |b|.
solution Takea = −3andb = 1. Thena < b but |a| = 3> 1= |b|.
2. Whichnumberssatisfy|a| = a?Whichsatisfy|a| = −a?What about |−a| = a?
solution Thenumbersa ≥ 0satisfy|a| = a and| −a| = a. Thenumbersa ≤ 0satisfy|a| = −a.
3. Giveanexampleof numbersa andb suchthat |a +b| < |a| +|b|.
solution Takea = −3andb = 1. Then
|a +b| = | −3+1| = | −2| = 2, but |a| +|b| = | −3| +|1| = 3+1= 4.
Thus, |a +b| < |a| +|b|.
4. What arethecoordinatesof thepoint lyingat theintersectionof thelinesx = 9andy = −4?
solution Thepoint (9, −4) liesat theintersectionof thelinesx = 9andy = −4.
5. Inwhichquadrant dothefollowingpointslie?
(a) (1, 4) (b) (−3, 2) (c) (4, −3) (d) (−4, −1)
solution
(a) Becauseboththex- andy-coordinatesof thepoint (1, 4) arepositive, thepoint (1, 4) liesintheﬁrst quadrant.
(b) Becausethex-coordinateof thepoint (−3, 2) is negativebut they-coordinateis positive, thepoint (−3, 2) lies in
thesecondquadrant.
(c) Becausethex-coordinateof thepoint (4, −3) is positivebut they-coordinateis negative, thepoint (4, −3) lies in
thefourthquadrant.
(d) Becauseboththex- andy-coordinatesof thepoint(−4, −1) arenegative, thepoint(−4, −1) liesinthethirdquadrant.
6. What istheradiusof thecirclewithequation(x −9)
2
+(y −9)
2
= 9?
solution Thecirclewithequation(x −9)
2
+(y −9)
2
= 9hasradius3.
7. Theequationf (x) = 5hasasolutionif (chooseone):
(a) 5belongstothedomainof f .
(b) 5belongstotherangeof f .
solution Thecorrect responseis(b): theequationf (x) = 5hasasolutionif 5belongstotherangeof f .
8. What kindof symmetrydoesthegraphhaveif f (−x) = −f (x)?
solution If f (−x) = −f (x), thenthegraphof f issymmetricwithrespect totheorigin.
Exercises
1. Useacalculator toﬁndarational number r suchthat |r −π
2
| < 10
−4
.
solution r must satisfy π
2
− 10
−4
< r < π
2
+ 10
−4
, or 9.869504 < r < 9.869705. r = 9.8696 =
12337
1250
would
beonesuchnumber.
Whichof (a)–(f) aretruefor a = −3andb = 2?
(a) a < b (b) |a| < |b| (c) ab > 0
(d) 3a < 3b (e) −4a < −4b (f)
1
a
<
1
b
In Exercises 3–8, express the interval in terms of an inequality involving absolute value.
3. [−2, 2]
solution |x| ≤ 2
(−4, 4)
5. (0, 4)
solution Themidpoint of theinterval isc = (0+ 4)/2 = 2, andtheradiusisr = (4− 0)/2 = 2; therefore, (0, 4)
canbeexpressedas|x −2| < 2.
[−4, 0]
7. [1, 5]
solution Themidpoint of theinterval is c = (1+ 5)/2 = 3, andtheradius is r = (5− 1)/2 = 2; therefore, the
interval [1, 5] canbeexpressedas|x −3| ≤ 2.
(−2, 8) 1
June 7, 2011 LTSV SSM Second Pass
2 C HA P T E R 1 PRECALCULUS REVIEW
In Exercises 9–12, write the inequality in the form a < x < b.
9. |x| < 8
solution −8< x < 8
|x −12| < 8
11. |2x +1| < 5
solution −5< 2x +1< 5so−6< 2x < 4and−3< x < 2
|3x −4| < 2
In Exercises 13–18, express the set of numbers x satisfying the given condition as an interval.
13. |x| < 4
solution (−4, 4)
|x| ≤ 9
15. |x −4| < 2
solution Theexpression|x −4| < 2isequivalent to−2< x −4< 2. Therefore, 2< x < 6, whichrepresentsthe
interval (2, 6).
|x +7| < 2
17. |4x −1| ≤ 8
solution Theexpression|4x −1| ≤ 8isequivalent to−8≤ 4x −1≤ 8or −7≤ 4x ≤ 9. Therefore, −
7
4
≤ x ≤
9
4
,
whichrepresentstheinterval [−
7
4
,
9
4
].
|3x +5| < 1
In Exercises 19–22, describe the set as a union of ﬁnite or inﬁnite intervals.
19. {x : |x −4| > 2}
solution x −4> 2or x −4< −2⇒ x > 6or x < 2⇒ (−∞, 2) ∪ (6, ∞)
{x : |2x +4| > 3}
21. {x : |x
2
−1| > 2}
solution x
2
− 1> 2or x
2
− 1< −2⇒ x
2
> 3or x
2
< −1(thiswill never happen) ⇒ x >

3or x < −

3⇒
(−∞, −

3) ∪ (

3, ∞).
{x : |x
2
+2x| > 2}
23. Match(a)–(f) with(i)–(vi).
(a) a > 3 (b) |a −5| <
1
3
(c)
¸
¸
¸
¸
a −
1
3
¸
¸
¸
¸
< 5 (d) |a| > 5
(e) |a −4| < 3 (f) 1≤ a ≤ 5
(i) a liestotheright of 3.
(ii) a liesbetween1and7.
(iii) Thedistancefroma to5islessthan
1
3
.
(iv) Thedistancefroma to3isat most 2.
(v) a islessthan5unitsfrom
1
3
.
(vi) a lieseither totheleft of −5or totheright of 5.
solution
(a) Onthenumber line, numbersgreater than3appear totheright; hence, a > 3isequivalenttothenumberstotheright
of 3: (i).
(b) |a −5| measuresthedistancefroma to5; hence, |a −5| <
1
3
issatisﬁedbythosenumberslessthan
1
3
of aunitfrom
5: (iii).
(c) |a −
1
3
| measuresthedistancefroma to
1
3
; hence, |a −
1
3
| < 5issatisﬁedbythosenumberslessthan5unitsfrom
1
3
: (v).
(d) Theinequality|a| > 5isequivalenttoa > 5or a < −5; thatis, either a liestotherightof 5or totheleftof −5: (vi).
(e) Theinterval describedbytheinequality|a − 4| < 3hasacenter at 4andaradiusof 3; that is, theinterval consists
of thosenumbersbetween1and7: (ii).
(f) Theinterval describedbytheinequality1< x < 5hasacenter at 3andaradiusof 2; that is, theinterval consistsof
thosenumberslessthan2unitsfrom3: (iv).
Describe
_
x :
x
x +1
< 0
_
asaninterval.
25. Describe{x : x
2
+2x < 3} asaninterval. Hint: Plot y = x
2
+2x −3.
solution Theinequalityx
2
+2x < 3isequivalent tox
2
+2x −3< 0. Intheﬁgurebelow, weseethat thegraphof
y = x
2
+ 2x − 3falls belowthex-axis for −3 < x < 1. Thus, theset {x : x
2
+ 2x < 3} corresponds totheinterval
−3< x < 1.
June 7, 2011 LTSV SSM Second Pass
S E C T I ON 1.1 Real Numbers, Functions, and Graphs 3
−4 −3 −2
−2
2
4
6
8
10
y
x
1 2
y = x
2
+ 2x − 3
Describetheset of real numberssatisfying|x −3| = |x −2| +1asahalf-inﬁniteinterval.
27. Showthat if a > b, thenb
−1
> a
−1
, providedthat a andb havethesamesign. What happensif a > 0andb < 0?
solution Case1a: If a andb arebothpositive, thena > b ⇒ 1>
b
a

1
b
>
1
a
.
Case1b: If a andb arebothnegative, thena > b ⇒ 1<
b
a
(sincea isnegative) ⇒
1
b
>
1
a
(again, sinceb isnegative).
Case2: If a > 0andb < 0, then
1
a
> 0and
1
b
< 0so
1
b
<
1
a
. (SeeExercise2f for anexampleof this).
Whichx satisfyboth|x −3| < 2and|x −5| < 1?
29. Showthat if |a −5| <
1
2
and|b −8| <
1
2
, then|(a +b) −13| < 1. Hint: Usethetriangleinequality.
solution
|a +b −13| = |(a −5) +(b −8)|
≤ |a −5| +|b −8| (bythetriangleinequality)
<
1
2
+
1
2
= 1.
Supposethat |x −4| ≤ 1.
(a) What isthemaximumpossiblevalueof |x +4|?
(b) Showthat |x
2
−16| ≤ 9.
31. Supposethat |a −6| ≤ 2and|b| ≤ 3.
(a) What isthelargest possiblevalueof |a +b|?
(b) What isthesmallest possiblevalueof |a +b|?
solution |a −6| ≤ 2guaranteesthat4≤ a ≤ 8,while|b| ≤ 3guaranteesthat−3≤ b ≤ 3.Therefore1≤ a +b ≤ 11.
It followsthat
(a) thelargest possiblevalueof |a +b| is11; and
(b) thesmallest possiblevalueof |a +b| is1.
Provethat |x| −|y| ≤ |x −y|. Hint: Applythetriangleinequalitytoy andx −y.
33. Expressr
1
= 0.27asafraction. Hint: 100r
1
−r
1
isaninteger. Thenexpressr
2
= 0.2666. . . asafraction.
solution Let r
1
= 0.27. Weobservethat 100r
1
= 27.27. Therefore, 100r
1
−r
1
= 27.27−0.27= 27and
r
1
=
27
99
=
3
11
.
Now, let r
2
= 0.2666. Then10r
2
= 2.666and100r
2
= 26.666. Therefore, 100r
2
−10r
2
= 26.666−2.666= 24and
r
2
=
24
90
=
4
15
.
Represent 1/7and4/27asrepeatingdecimals.
35. Thetext states: If the decimal expansions of numbers a and b agree to k places, then |a −b| ≤ 10
−k
. Showthat the
converseisfalse: For all k therearenumbersa andb whosedecimal expansionsdo not agree at all but |a −b| ≤ 10
−k
.
solution Let a = 1andb = 0.9(seethediscussionbeforeExample1). Thedecimal expansions of a andb donot
agree, but |1−0.9| < 10
−k
for all k.
Plot eachpair of pointsandcomputethedistancebetweenthem:
(a) (1, 4) and(3, 2) (b) (2, 1) and(2, 4)
(c) (0, 0) and(−2, 3) (d) (−3, −3) and(−2, 3)
37. Findtheequationof thecirclewithcenter (2, 4):
(a) withradiusr = 3.
(b) that passesthrough(1, −1).
solution
(a) Theequationof theindicatedcircleis(x −2)
2
+(y −4)
2
= 3
2
= 9.
(b) First determinetheradiusasthedistancefromthecenter totheindicatedpoint onthecircle:
r =
_
(2−1)
2
+(4−(−1))
2
=

26.
Thus, theequationof thecircleis(x −2)
2
+(y −4)
2
= 26.
Findall pointswithinteger coordinateslocatedat adistance5fromtheorigin. Thenﬁndall pointswithinteger
coordinateslocatedat adistance5from(2, 3).
39. Determinethedomainandrangeof thefunction
f : {r, s, t, u} → {A, B, C, D, E}
deﬁnedbyf (r) = A, f (s) = B, f (t ) = B, f (u) = E.
solution Thedomainistheset D = {r, s, t, u}; therangeistheset R = {A, B, E}.
June 7, 2011 LTSV SSM Second Pass
4 C HA P T E R 1 PRECALCULUS REVIEW
Giveanexampleof afunctionwhosedomainD hasthreeelementsandwhoserangeR hastwoelements. Doesa
functionexist whosedomainD hastwoelementsandwhoserangeR hasthreeelements?
In Exercises 41–48, ﬁnd the domain and range of the function.
41. f (x) = −x
solution D : all reals; R : all reals
g(t ) = t
4
43. f (x) = x
3
solution D : all reals; R : all reals
g(t ) =

2−t
45. f (x) = |x|
solution D : all reals; R : {y : y ≥ 0}
h(s) =
1
s
47. f (x) =
1
x
2
solution D : {x : x = 0}; R : {y : y > 0}
g(t ) = cos
1
t
In Exercises 49–52, determine where f (x) is increasing.
49. f (x) = |x +1|
solution A graphof thefunctiony = |x +1| isshownbelow. Fromthegraph, weseethat thefunctionisincreasing
ontheinterval (−1, ∞).
x
−3 −2 −1
1
2
1
y
f (x) = x
3
51. f (x) = x
4
solution A graphof thefunctiony = x
4
isshownbelow. Fromthegraph, weseethat thefunctionisincreasingon
theinterval (0, ∞).
x
−2 −1 1 2
12
4
8
y
f (x) =
1
x
4
+x
2
+1
In Exercises 53–58, ﬁnd the zeros of f (x) and sketch its graph by plotting points. Use symmetry and increase/decrease
information where appropriate.
53. f (x) = x
2
−4
solution Zeros: ±2
Increasing: x > 0
Decreasing: x < 0
Symmetry: f (−x) = f (x) (evenfunction). So, y-axissymmetry.
2
−2
−4
4
y
x
−2 −1 1 2
f (x) = 2x
2
−4
55. f (x) = x
3
−4x
solution Zeros: 0, ±2; Symmetry: f (−x) = −f (x) (oddfunction). Sooriginsymmetry.
5
−5
−10
10
y
x
−2 −1 1 2
June 7, 2011 LTSV SSM Second Pass
S E C T I ON 1.1 Real Numbers, Functions, and Graphs 5
f (x) = x
3
57. f (x) = 2−x
3
solution Thisisanx-axisreﬂectionof x
3
translatedup2units. Thereisonezeroat x =
3

2.
10
−10
−20
20
y
x
−2 −1 1 2
f (x) =
1
(x −1)
2
+1
59. Whichof thecurvesinFigure26isthegraphof afunction?
(A)
x
y
(B)
x
y
(C)
x
y
(D)
x
y
FIGURE 26
solution (B) isthegraphof afunction. (A), (C), and(D) all fail thevertical linetest.
Determinewhether thefunctioniseven, odd, or neither.
(a) f (x) = x
5
(b) g(t ) = t
3
−t
2
(c) F(t ) =
1
t
4
+t
2
61. Determinewhether thefunctioniseven, odd, or neither.
(a) f (t ) =
1
t
4
+t +1

1
t
4
−t +1
(b) g(t ) = 2
t
−2
−t
(c) G(θ) = sinθ +cosθ (d) H(θ) = sin(θ
2
)
solution
(a) Thisfunctionisoddbecause
f (−t ) =
1
(−t )
4
+(−t ) +1

1
(−t )
4
−(−t ) +1
=
1
t
4
−t +1

1
t
4
+t +1
= −f (t ).
(b) g(−t ) = 2
−t
−2
−(−t )
= 2
−t
−2
t
= −g(t ), sothisfunctionisodd.
(c) G(−θ) = sin(−θ) + cos(−θ) = −sinθ + cosθ which is equal to neither G(θ) nor −G(θ), so this function is
neither oddnor even.
(d) H(−θ) = sin((−θ)
2
) = sin(θ
2
) = H(θ), sothisfunctioniseven.
Writef (x) = 2x
4
−5x
3
+12x
2
−3x +4asthesumof anevenandanoddfunction.
63. Determinetheinterval onwhichf (x) =
1
x −4
isincreasingor decreasing.
solution Agraphof thefunctionisshownbelow. Fromthisgraphwecanseethatf (x) isdecreasingon(−∞, 4) and
alsodecreasingon(4, ∞).
−2
−2
0
−4
−6
2
4
6
2 4 6 8 10
x
y
June 7, 2011 LTSV SSM Second Pass
6 C HA P T E R 1 PRECALCULUS REVIEW
Statewhether thefunctionisincreasing, decreasing, or neither.
(a) Surfaceareaof asphereasafunctionof itsradius
(b) Temperatureat apoint ontheequator asafunctionof time
(c) Priceof anairlineticket asafunctionof thepriceof oil
(d) Pressureof thegasinapistonasafunctionof volume
In Exercises 65–70, let f (x) be the function shown in Figure 27.
1 2 3 4
0
1
2
3
4
x
y
FIGURE 27
65. Findthedomainandrangeof f (x)?
solution D : [0, 4]; R : [0, 4]
Sketchthegraphsof f (x +2) andf (x) +2.
67. Sketchthegraphsof f (2x), f
_
1
2
x
_
, and2f (x).
solution Thegraphof y = f (2x) isobtainedbycompressingthegraphof y = f (x) horizontallybyafactor of 2(see
thegraphbelowontheleft). Thegraphof y = f (
1
2
x) isobtainedbystretchingthegraphof y = f (x) horizontallybya
factor of 2(seethegraphbelowinthemiddle). Thegraphof y = 2f (x) isobtainedbystretchingthegraphof y = f (x)
verticallybyafactor of 2(seethegraphbelowontheright).
y
x
1
2
3
4
1 2 3 4
f (2x)
y
x
1
2
3
4
2 4 6 8
f (x/2)
y
x
2
4
6
8
1 2 3 4
2f (x)
Sketchthegraphsof f (−x) and−f (−x).
69. Extendthegraphof f (x) to[−4, 4] sothat it isanevenfunction.
solution Tocontinuethegraphof f (x) totheinterval [−4, 4] asanevenfunction, reﬂect thegraphof f (x) across
they-axis(seethegraphbelow).
−2 −4
x
2 4
y
1
2
3
4
Extendthegraphof f (x) to[−4, 4] sothat it isanoddfunction.
71. Supposethat f (x) hasdomain[4, 8] andrange[2, 6]. Findthedomainandrangeof:
(a) f (x) +3 (b) f (x +3)
(c) f (3x) (d) 3f (x)
solution
(a) f (x) + 3 is obtained by shifting f (x) upward threeunits. Therefore, thedomain remains [4, 8], whiletherange
becomes[5, 9].
(b) f (x +3) isobtainedbyshiftingf (x) leftthreeunits. Therefore, thedomainbecomes[1, 5], whiletherangeremains
[2, 6].
(c) f (3x) is obtained by compressingf (x) horizontally by afactor of three. Therefore, thedomain becomes [
4
3
,
8
3
],
whiletherangeremains[2, 6].
(d) 3f (x) isobtainedbystretchingf (x) verticallybyafactor of three. Therefore, thedomainremains[4, 8], whilethe
rangebecomes[6, 18].
Let f (x) = x
2
. Sketchthegraphover [−2, 2] of:
(a) f (x +1) (b) f (x) +1
(c) f (5x) (d) 5f (x)
73. Supposethat thegraphof f (x) = sinx is compressedhorizontally by afactor of 2andthenshifted5units to the
right.
(a) What istheequationfor thenewgraph?
(b) What istheequationif youﬁrst shift by5andthencompressby2?
(c) Verifyyour answersbyplottingyour equations.
June 7, 2011 LTSV SSM Second Pass
S E C T I ON 1.1 Real Numbers, Functions, and Graphs 7
solution
(a) Let f (x) = sinx. After compressingthegraphof f horizontally by afactor of 2, weobtainthefunctiong(x) =
f (2x) = sin2x. Shiftingthegraph5unitstotheright thenyields
h(x) = g(x −5) = sin2(x −5) = sin(2x −10).
(b) Letf (x) = sinx. After shiftingthegraph5unitstotheright, weobtainthefunctiong(x) = f (x −5) = sin(x −5).
Compressingthegraphhorizontallybyafactor of 2thenyields
h(x) = g(2x) = sin(2x −5).
(c) The ﬁgure below at the top left shows the graphs of y = sinx (the dashed curve), the sine graph compressed
horizontallybyafactor of 2(thedash, doubledotcurve) andthenshiftedright5units(thesolidcurve). Comparethislast
graphwiththegraphof y = sin(2x −10) shownat thebottomleft.
Theﬁgurebelowat thetopright showsthegraphsof y = sinx (thedashedcurve), thesinegraphshiftedtotheright
5units(thedash, doubledotcurve) andthencompressedhorizontallybyafactor of 2(thesolidcurve). Comparethislast
graphwiththegraphof y = sin(2x −5) shownat thebottomright.
−1
1
y
x
−6 −4 −2 6 4 2
−1
1
y
x
−6 −4 −2 6 4 2
−1
1
y
x
−6 −4 −2 6 4 2
−1
1
y
x
−6 −4 −2 6 4 2
Figure28showsthegraphof f (x) = |x| +1. Matchthefunctions(a)–(e) withtheir graphs(i)–(v).
(a) f (x −1) (b) −f (x) (c) −f (x) +2
(d) f (x −1) −2 (e) f (x +1)
75. Sketchthegraphof f (2x) andf
_
1
2
x
_
, wheref (x) = |x| +1(Figure28).
solution Thegraphof y = f (2x) is obtainedby compressingthegraphof y = f (x) horizontally by afactor of 2
(seethegraphbelowontheleft). Thegraphof y = f (
1
2
x) isobtainedbystretchingthegraphof y = f (x) horizontally
byafactor of 2(seethegraphbelowontheright).
x
−1
2
4
6
−2 −3 1 2 3
y
f (2x)
x
−1
2
4
6
−2 −3 1 2 3
y
f (x/2)
Findthefunctionf (x) whosegraphisobtainedbyshiftingtheparabolay = x
2
threeunitstotheright andfour
unitsdown, asinFigure29.
77. Deﬁnef (x) tobethelarger of x and2−x. Sketchthegraphof f (x). Whatareitsdomainandrange?Expressf (x)
intermsof theabsolutevaluefunction.
solution
x
−1
1
2
1 2 3
y
Thegraphof y = f (x) isshownabove. Clearly, thedomainof f isthesetof all real numberswhiletherangeis{y | y ≥ 1}.
Noticethegraphhas thestandardV-shapeassociatedwiththeabsolutevaluefunction, but thebaseof theV has been
translatedtothepoint (1, 1). Thus, f (x) = |x −1| +1.
For eachcurveinFigure30, statewhether it issymmetricwithrespect tothey-axis, theorigin, both, or neither.
79. Showthat thesumof twoevenfunctionsisevenandthesumof twooddfunctionsisodd.
solution Even: (f +g)(−x) = f (−x) +g(−x)
even
= f (x) +g(x) = (f +g)(x)
Odd: (f +g)(−x) = f (−x) +g(−x)
odd
= −f (x) +−g(x) = −(f +g)(x)
June 7, 2011 LTSV SSM Second Pass
8 C HA P T E R 1 PRECALCULUS REVIEW
Supposethat f (x) andg(x) arebothodd. Whichof thefollowingfunctionsareeven?Whichareodd?
(a) f (x)g(x) (b) f (x)
3
(c) f (x) −g(x) (d)
f (x)
g(x)
81. Provethat theonlyfunctionwhosegraphissymmetricwithrespect toboththey-axisandtheoriginisthefunction
f (x) = 0.
solution Supposef issymmetricwithrespecttothey-axis. Thenf (−x) = f (x). If f isalsosymmetricwithrespect
totheorigin, thenf (−x) = −f (x). Thusf (x) = −f (x) or 2f (x) = 0. Finally, f (x) = 0.
Further Insights and Challenges
Provethetriangleinequalitybyaddingthetwoinequalities
−|a| ≤ a ≤ |a|, −|b| ≤ b ≤ |b|
83. Showthat afractionr = a/b inlowest termshasaﬁnite decimal expansionif andonlyif
b = 2
n
5
m
for somen, m ≥ 0.
Hint: Observethat r hasaﬁnitedecimal expansionwhen10
N
r isaninteger for someN ≥ 0(andhenceb divides10
N
).
solution Supposer hasaﬁnitedecimal expansion. Thenthereexistsaninteger N ≥ 0suchthat 10
N
r isaninteger,
call it k. Thus, r = k/10
N
. Becausetheonlyprimefactorsof 10are2and5, it followsthat whenr iswritteninlowest
terms, itsdenominator must beof theform2
n
5
m
for someintegersn, m ≥ 0.
Conversely, suppose r = a/b in lowest with b = 2
n
5
m
for some integers n, m ≥ 0. Then r =
a
b
=
a
2
n
5
m
or
2
n
5
m
r = a. If m ≥ n, then2
m
5
m
r = a2
m−n
or r =
a2
m−n
10
m
andthus r has aﬁnitedecimal expansion(less thanor
equal tom terms, tobeprecise). Ontheother hand, if n > m, then2
n
5
n
r = a5
n−m
or r =
a5
n−m
10
n
andonceagainr has
aﬁnitedecimal expansion.
Let p = p
1
. . . p
s
beaninteger withdigitsp
1
, . . . , p
s
. Showthat
p
10
s
−1
= 0.p
1
. . . p
s
Usethistoﬁndthedecimal expansionof r =
2
11
. Notethat
r =
2
11
=
18
10
2
−1
85. A functionf (x) issymmetricwithrespect tothevertical linex = a if f (a −x) = f (a +x).
(a) Drawthegraphof afunctionthat issymmetricwithrespect tox = 2.
(b) Showthat if f (x) issymmetricwithrespect tox = a, theng(x) = f (x +a) iseven.
solution
(a) Therearemanypossibilities, oneof whichis
x
−1
1
2
5 4 3 2 1
y
y = |x − 2|
(b) Let g(x) = f (x +a). Then
g(−x) = f (−x +a) = f (a −x)
= f (a +x) symmetrywithrespect tox = a
= g(x)
Thus, g(x) iseven.
Formulateaconditionfor f (x) tobesymmetricwithrespect tothepoint (a, 0) onthex-axis.
1.2 Linear and Quadratic Functions
Preliminary Questions
1. What istheslopeof theliney = −4x −9?
solution Theslopeof theliney = −4x −9is−4, givenbythecoefﬁcient of x.
2. Arethelinesy = 2x +1andy = −2x −4perpendicular?
solution Theslopesof perpendicular linesarenegativereciprocalsof oneanother. Becausetheslopeof y = 2x +1
is2andtheslopeof y = −2x −4is−2, thesetwolinesarenot perpendicular.
3. Whenisthelineax +by = c parallel tothey-axis?Tothex-axis?
solution Thelineax +by = c will beparallel tothey-axiswhenb = 0andparallel tothex-axiswhena = 0.
4. Supposey = 3x +2. What isy if x increasesby3?
solution Becausey = 3x +2isalinear functionwithslope3, increasingx by3will leadtoy = 3(3) = 9.
June 7, 2011 LTSV SSM Second Pass
S E C T I ON 1.2 Linear and Quadratic Functions 9
5. What istheminimumof f (x) = (x +3)
2
−4?
solution Because(x +3)
2
≥ 0, it followsthat(x +3)
2
−4≥ −4. Thus, theminimumvalueof (x +3)
2
−4is−4.
6. What istheresult of completingthesquarefor f (x) = x
2
+1?
solution Becausethereisnox terminx
2
+1, completingthesquareonthisexpressionleadsto(x −0)
2
+1.
Exercises
In Exercises 1–4, ﬁnd the slope, the y-intercept, and the x-intercept of the line with the given equation.
1. y = 3x +12
solution Becausetheequationof thelineis giveninslope-intercept form, theslopeis thecoefﬁcient of x andthe
y-intercept istheconstant term: that is, m = 3andthey-intercept is12. Todeterminethex-intercept, substitutey = 0
andthensolvefor x: 0= 3x +12or x = −4.
y = 4−x
3. 4x +9y = 3
solution Todeterminetheslopeandy-intercept, weﬁrst solvetheequationfor y toobtaintheslope-intercept form.
This yields y = −
4
9
x +
1
3
. Fromhere, weseethat theslopeis m = −
4
9
and they-intercept is
1
3
. To determinethe
x-intercept, substitutey = 0andsolvefor x: 4x = 3or x =
3
4
.
y −3=
1
2
(x −6)
In Exercises 5–8, ﬁnd the slope of the line.
5. y = 3x +2
solution m = 3
y = 3(x −9) +2
7. 3x +4y = 12
solution First solvetheequationfor y toobtaintheslope-intercept form. Thisyieldsy = −
3
4
x +3. Theslopeof the
lineisthereforem = −
3
4
.
3x +4y = −8
In Exercises 9–20, ﬁnd the equation of the line with the given description.
9. Slope3, y-intercept 8
solution Usingtheslope-intercept formfor theequationof aline, wehavey = 3x +8.
Slope−2, y-intercept 3
11. Slope3, passesthrough(7, 9)
solution Usingthepoint-slopeformfor theequationof aline, wehavey −9= 3(x −7) or y = 3x −12.
Slope−5, passesthrough(0, 0)
13. Horizontal, passesthrough(0, −2)
solution Ahorizontal linehasaslopeof 0. Usingthepoint-slopeformfor theequationof aline, wehavey −(−2) =
0(x −0) or y = −2.
Passesthrough(−1, 4) and(2, 7)
15. Parallel toy = 3x −4, passesthrough(1, 1)
solution Becausetheequationy = 3x −4isinslope-intercept form, wecanreadilyidentifythat it hasaslopeof 3.
Parallel lineshavethesameslope, sotheslopeof therequestedlineisalso3. Usingthepoint-slopeformfor theequation
of aline, wehavey −1= 3(x −1) or y = 3x −2.
Passesthrough(1, 4) and(12, −3)
17. Perpendicular to3x +5y = 9, passesthrough(2, 3)
solution Westart bysolvingtheequation3x +5y = 9for y toobtaintheslope-intercept formfor theequationof a
line. Thisyields
y = −
3
5
x +
9
5
,
fromwhichweidentifytheslopeas−
3
5
. Perpendicular lineshaveslopesthat arenegativereciprocalsof oneanother, so
theslopeof thedesiredlineism

=
5
3
. Usingthepoint-slopeformfor theequationof aline, wehavey −3=
5
3
(x −2)
or y =
5
3
x −
1
3
.
Vertical, passesthrough(−4, 9)
19. Horizontal, passesthrough(8, 4)
solution Ahorizontal linehasslope0. Usingthepointslopeformfortheequationof aline, wehavey −4= 0(x −8)
or y = 4.
Slope3, x-intercept 6
June 7, 2011 LTSV SSM Second Pass
10 C HA P T E R 1 PRECALCULUS REVIEW
21. Findtheequationof theperpendicularbisectorof thesegmentjoining(1, 2) and(5, 4) (Figure11). Hint: Themidpoint
Qof thesegment joining(a, b) and(c, d) is
_
a +c
2
,
b +d
2
_
.
Q
(1, 2)
(5, 4)
Perpendicular
bisector
x
y
FIGURE 11
solution Theslopeof thesegment joining(1, 2) and(5, 4) is
m =
4−2
5−1
=
1
2
andthemidpoint of thesegment (Figure11) is
midpoint =
_
1+5
2
,
2+4
2
_
= (3, 3)
Theperpendicular bisector hasslope−1/m = −2andpassesthrough(3, 3), soitsequationis: y − 3 = −2(x − 3) or
y = −2x +9.
Intercept-Intercept Form Showthat if a, b = 0, thenthelinewithx-intercept x = a andy-intercept y = b
hasequation(Figure12)
x
a
+
y
b
= 1
23. Findanequationof thelinewithx-intercept x = 4andy-intercept y = 3.
solution FromExercise22,
x
4
+
y
3
= 1or 3x +4y = 12.
Findy suchthat (3, y) liesonthelineof slopem = 2through(1, 4).
25. Determinewhether thereexistsaconstant c suchthat thelinex +cy = 1:
(a) Hasslope4 (b) Passesthrough(3, 1)
(c) Ishorizontal (d) Isvertical
solution
(a) Rewritingtheequationof thelineinslope-intercept formgivesy = −
x
c
+
1
c
. Tohaveslope4requires−
1
c
= 4or
c = −
1
4
.
(b) Substitutingx = 3andy = 1intotheequationof thelinegives3+c = 1or c = −2.
(c) From(a), weknowtheslopeof thelineis−
1
c
. Thereisnovaluefor c that will makethisslopeequal to0.
(d) Withc = 0, theequationbecomesx = 1. Thisistheequationof avertical line.
Assumethat thenumber N of concert ticketsthat canbesoldat apriceof P dollarsper ticket isalinear function
N(P) for 10≤ P ≤ 40. DetermineN(P) (calledthedemandfunction) if N(10) = 500andN(40) = 0. What isthe
decreaseN inthenumber of ticketssoldif thepriceisincreasedbyP = 5dollars?
27. Materialsexpandwhenheated. Consider ametal rodof lengthL
0
at temperatureT
0
. If thetemperatureischanged
by anamount T , thentherod’slengthchangesby L = αL
0
T , whereα isthethermal expansioncoefﬁcient. For
steel, α = 1.24×10
−5◦
C
−1
.
(a) A steel rodhaslengthL
0
= 40cmat T
0
= 40

C. Finditslengthat T = 90

C.
(b) Finditslengthat T = 50

C if itslengthat T
0
= 100

C is65cm.
(c) ExpresslengthL asafunctionof T if L
0
= 65cmat T
0
= 100

C.
solution
(a) WithT = 90

C andT
0
= 40

C, T = 50

C. Therefore,
L = αL
0
T = (1.24×10
−5
)(40)(50) = 0.0248 and L = L
0
+L = 40.0248cm.
(b) WithT = 50

C andT
0
= 100

C, T = −50

C. Therefore,
L = αL
0
T = (1.24×10
−5
)(65)(−50) = −0.0403 and L = L
0
+L = 64.9597cm.
(c) L = L
0
+L = L
0
+αL
0
T = L
0
(1+αT ) = 65(1+α(T −100))
Dothepoints(0.5, 1), (1, 1.2), (2, 2) lieonaline?
29. Findb suchthat (2, −1), (3, 2), and(b, 5) lieonaline.
solution Theslopeof thelinedeterminedbythepoints(2, −1) and(3, 2) is
2−(−1)
3−2
= 3.
Tolieonthesameline, theslopebetween(3, 2) and(b, 5) must alsobe3. Thus, werequire
5−2
b −3
=
3
b −3
= 3,
or b = 4.
June 7, 2011 LTSV SSM Second Pass
S E C T I ON 1.2 Linear and Quadratic Functions 11
Findanexpressionfor thevelocityv asalinear functionof t that matchesthefollowingdata.
t (s) 0 2 4 6
v (m/s) 39.2 58.6 78 97.4
31. TheperiodT of apendulumismeasuredfor pendulumsof several different lengthsL. Basedonthefollowingdata,
doesT appear tobealinear functionof L?
L (cm) 20 30 40 50
T (s) 0.9 1.1 1.27 1.42
solution Examinetheslopebetweenconsecutivedatapoints. Theﬁrst pair of datapointsyieldsaslopeof
1.1−0.9
30−20
= 0.02,
whilethesecondpair of datapointsyieldsaslopeof
1.27−1.1
40−30
= 0.017,
andthelast pair of datapointsyieldsaslopeof
1.42−1.27
50−40
= 0.015
Becausethethreeslopesarenot equal, T doesnot appear tobealinear functionof L.
Showthat f (x) islinear of slopem if andonlyif
f (x +h) −f (x) = mh (for all x andh)
33. Findtherootsof thequadraticpolynomials:
(a) 4x
2
−3x −1 (b) x
2
−2x −1
solution
(a) x =

9−4(4)(−1)
2(4)
=

25
8
= 1or −
1
4
(b) x =

4−(4)(1)(−1)
2
=

8
2
= 1±

2
In Exercises 34–41, complete the square and ﬁnd the minimum or maximum value of the quadratic function.
y = x
2
+2x +5
35. y = x
2
−6x +9
solution y = (x −3)
2
; therefore, theminimumvalueof thequadraticpolynomial is0, andthisoccursat x = 3.
y = −9x
2
+x
37. y = x
2
+6x +2
solution y = x
2
+ 6x + 9− 9+ 2 = (x + 3)
2
− 7; therefore, theminimumvalueof thequadratic polynomial is
−7, andthisoccursat x = −3.
y = 2x
2
−4x −7
39. y = −4x
2
+3x +8
solution y = −4x
2
+3x +8= −4(x
2

3
4
x +
9
64
) +8+
9
16
= −4(x −
3
8
)
2
+
137
16
; therefore, themaximumvalue
of thequadraticpolynomial is
137
16
, andthisoccursat x =
3
8
.
y = 3x
2
+12x −5
41. y = 4x −12x
2
solution y = −12(x
2

x
3
) = −12(x
2

x
3
+
1
36
) +
1
3
= −12(x −
1
6
)
2
+
1
3
; therefore, themaximumvalueof the
quadraticpolynomial is
1
3
, andthisoccursat x =
1
6
.
Sketchthegraphof y = x
2
−6x +8byplottingtherootsandtheminimumpoint.
43. Sketchthegraphof y = x
2
+4x +6byplottingtheminimumpoint, they-intercept, andoneother point.
solution y = x
2
+ 4x + 4− 4+ 6= (x + 2)
2
+ 2sotheminimumoccursat (−2, 2). If x = 0, theny = 6andif
x = −4, y = 6. Thisisthegraphof x
2
movedleft 2unitsandup2units.
−4 −3 −2 −1
2
4
6
8
10
y
x
If thealleles A andB of thecystic ﬁbrosis geneoccur inapopulationwithfrequencies p and1− p (wherep
isafractionbetween0and1), thenthefrequencyof heterozygouscarriers(carrierswithbothalleles) is2p(1−p).
Whichvalueof p givesthelargest frequencyof heterozygouscarriers?
June 7, 2011 LTSV SSM Second Pass
12 C HA P T E R 1 PRECALCULUS REVIEW
45. For whichvaluesof c doesf (x) = x
2
+cx +1haveadoubleroot? Noreal roots?
solution A doubleroot occurswhenc
2
−4(1)(1) = 0or c
2
= 4. Thus, c = ±2.
Therearenoreal rootswhenc
2
−4(1)(1) < 0or c
2
< 4. Thus, −2< c < 2.
Let f (x) be a quadratic function and c a constant. Which of the following statements is correct? Explain
graphically.
(a) Thereisauniquevalueof c suchthat y = f (x) −c hasadoubleroot.
(b) Thereisauniquevalueof c suchthat y = f (x −c) hasadoubleroot.
47. Provethat x +
1
x
≥ 2for all x > 0. Hint: Consider (x
1/2
−x
−1/2
)
2
.
solution Let x > 0. Then
_
x
1/2
−x
−1/2
_
2
= x −2+
1
x
.
Because(x
1/2
−x
−1/2
)
2
≥ 0, it followsthat
x −2+
1
x
≥ 0 or x +
1
x
≥ 2.
Let a, b > 0. Showthat thegeometric mean

ab isnot larger thanthearithmetic mean (a + b)/2. Hint: Usea
variationof thehint giveninExercise47.
49. If objects of weights x and w
1
are suspended fromthe balance in Figure 13(A), the cross-beamis horizontal if
bx = aw
1
. If thelengthsa andb areknown, wemayusethisequationtodetermineanunknownweightx byselectingw
1
suchthat thecross-beamishorizontal. If a andb arenot knownprecisely, wemight proceedasfollows. First balancex
byw
1
ontheleft asin(A). Thenswitchplacesandbalancex byw
2
ontheright asin(B). Theaverage ¯ x =
1
2
(w
1
+w
2
)
givesanestimatefor x. Showthat ¯ x isgreater thanor equal tothetrueweight x.
w
1
(A)
a
x
b
(B)
w
2
x
a b
FIGURE 13
solution First notebx = aw
1
andax = bw
2
. Thus,
¯ x =
1
2
(w
1
+w
2
)
=
1
2
_
bx
a
+
ax
b
_
=
x
2
_
b
a
+
a
b
_

x
2
(2) byExercise47
= x
Findnumbersx andy withsum10andproduct 24. Hint: Findaquadraticpolynomial satisﬁedbyx.
51. Findapair of numberswhosesumandproduct arebothequal to8.
solution Let x andy benumberswhosesumandproduct arebothequal to8. Thenx +y = 8andxy = 8. Fromthe
secondequation, y =
8
x
. Substitutingthisexpressionfor y intheﬁrst equationgivesx +
8
x
= 8or x
2
−8x +8= 0. By
thequadraticformula,
x =

64−32
2
= 4±2

2.
If x = 4+2

2, then
y =
8
4+2

2
=
8
4+2

2
·
4−2

2
4−2

2
= 4−2

2.
Ontheother hand, if x = 4−2

2, then
y =
8
4−2

2
=
8
4−2

2
·
4+2

2
4+2

2
= 4+2

2.
Thus, thetwonumbersare4+2

2and4−2

2.
Showthat theparabolay = x
2
consists of all points P suchthat d
1
= d
2
, whered
1
is thedistancefromP to
_
0,
1
4
_
andd
2
isthedistancefromP totheliney = −
1
4
(Figure14).
June 7, 2011 LTSV SSM Second Pass
S E C T I ON 1.3 The Basic Classes of Functions 13
Further Insights and Challenges
53. Showthat if f (x) andg(x) arelinear, thensoisf (x) +g(x). Isthesametrueof f (x)g(x)?
solution If f (x) = mx +b andg(x) = nx +d, then
f (x) +g(x) = mx +b +nx +d = (m+n)x +(b +d),
whichislinear. f (x)g(x) isnot generallylinear. Take, for example, f (x) = g(x) = x. Thenf (x)g(x) = x
2
.
Showthat if f (x) andg(x) arelinear functionssuchthat f (0) = g(0) andf (1) = g(1), thenf (x) = g(x).
55. Showthaty/x for thefunctionf (x) = x
2
over theinterval [x
1
, x
2
] isnotaconstant, butdependsontheinterval.
Determinetheexact dependenceof y/x onx
1
andx
2
.
solution For x
2
,
y
x
=
x
2
2
−x
2
1
x
2
−x
1
= x
2
+x
1
.
UseEq. (2) toderivethequadraticformulafor therootsof ax
2
+bx +c = 0.
57. Let a, c = 0. Showthat therootsof
ax
2
+bx +c = 0 and cx
2
+bx +a = 0
arereciprocalsof eachother.
solution Let r
1
andr
2
betheroots of ax
2
+ bx + c andr
3
andr
4
betheroots of cx
2
+ bx + a. Without loss of
generality, let
r
1
=
−b +
_
b
2
−4ac
2a

1
r
1
=
2a
−b +
_
b
2
−4ac
·
−b −
_
b
2
−4ac
−b −
_
b
2
−4ac
=
2a(−b −
_
b
2
−4ac)
b
2
−b
2
+4ac
=
−b −
_
b
2
−4ac
2c
= r
4
.
Similarly, youcanshow
1
r
2
= r
3
.
Show, bycompletingthesquare, that theparabola
y = ax
2
+bx +c
iscongruent toy = ax
2
byavertical andhorizontal translation.
59. ProveViète’s Formulas: Thequadratic polynomial withα andβ asrootsisx
2
+ bx + c, whereb = −α − β and
c = αβ.
solution If aquadraticpolynomial hasrootsα andβ, thenthepolynomial is
(x −α)(x −β) = x
2
−αx −βx +αβ = x
2
+(−α −β)x +αβ.
Thus, b = −α −β andc = αβ.
1.3 The Basic Classes of Functions
Preliminary Questions
1. Giveanexampleof arational function.
solution Oneexampleis
3x
2
−2
7x
3
+x −1
.
2. Is|x| apolynomial function?What about |x
2
+1|?
solution |x| isnot apolynomial; however, becausex
2
+1> 0for all x, it followsthat |x
2
+1| = x
2
+1, whichis
apolynomial.
3. Whatisunusual aboutthedomainof thecompositefunctionf ◦ g forthefunctionsf (x) = x
1/2
andg(x) = −1−|x|?
solution Recall that (f ◦ g)(x) = f (g(x)). Now, for any real number x, g(x) = −1− |x| ≤ −1< 0. Becausewe
cannot takethesquareroot of anegativenumber, it follows that f (g(x)) is not deﬁnedfor any real number. Inother
words, thedomainof f (g(x)) istheemptyset.
4. Isf (x) =
_
1
2
_
x
increasingor decreasing?
solution Thefunctionf (x) = (
1
2
)
x
is anexponential functionwithbaseb =
1
2
< 1. Therefore, f is adecreasing
function.
5. Giveanexampleof atranscendental function.
solution Onepossibilityisf (x) = e
x
−sinx.
June 7, 2011 LTSV SSM Second Pass
14 C HA P T E R 1 PRECALCULUS REVIEW
Exercises
In Exercises 1–12, determine the domain of the function.
1. f (x) = x
1/4
solution x ≥ 0
g(t ) = t
2/3
3. f (x) = x
3
+3x −4
solution All reals
h(z) = z
3
+z
−3
5. g(t ) =
1
t +2
solution t = −2
f (x) =
1
x
2
+4
7. G(u) =
1
u
2
−4
solution u = ±2
f (x) =

x
x
2
−9
9. f (x) = x
−4
+(x −1)
−3
solution x = 0, 1
F(s) = sin
_
s
s +1
_
11. g(y) = 10

y+y
−1
solution y > 0
f (x) =
x +x
−1
(x −3)(x +4)
In Exercises 13–24, identify each of the following functions as polynomial, rational, algebraic, or transcendental.
13. f (x) = 4x
3
+9x
2
−8
solution Polynomial
f (x) = x
−4
15. f (x) =

x
solution Algebraic
f (x) =
_
1−x
2 17. f (x) =
x
2
x +sinx
solution Transcendental
f (x) = 2
x
19. f (x) =
2x
3
+3x
9−7x
2
solution Rational
f (x) =
3x −9x
−1/2
9−7x
2
21. f (x) = sin(x
2
)
solution Transcendental
f (x) =
x

x +1
23. f (x) = x
2
+3x
−1
solution Rational
f (x) = sin(3
x
)
25. Isf (x) = 2
x
2
atranscendental function?
solution Yes.
Showthat f (x) = x
2
+3x
−1
andg(x) = 3x
3
−9x +x
−2
arerational functions—that is, quotientsof polyno-
mials.
In Exercises 27–34, calculate the composite functions f ◦ g and g ◦ f , and determine their domains.
27. f (x) =

x, g(x) = x +1
solution f (g(x)) =

x +1; D: x ≥ −1, g(f (x)) =

x +1; D: x ≥ 0
f (x) =
1
x
, g(x) = x
−4
29. f (x) = 2
x
, g(x) = x
2
solution f (g(x)) = 2
x
2
; D: R, g(f (x)) = (2
x
)
2
= 2
2x
; D: R
f (x) = |x|, g(θ) = sinθ
31. f (θ) = cosθ, g(x) = x
3
+x
2
solution f (g(x)) = cos(x
3
+x
2
); D: R, g(f (θ)) = cos
3
θ +cos
2
θ; D: R
f (x) =
1
x
2
+1
, g(x) = x
−2
33. f (t ) =
1

t
, g(t ) = −t
2
solution f (g(t )) =
1
_
−t
2
; D: Not validfor anyt , g(f (t )) = −
_
1

t
_
2
= −
1
t
; D: t > 0
f (t ) =

t , g(t ) = 1−t
3
June 7, 2011 LTSV SSM Second Pass
S E C T I ON 1.3 The Basic Classes of Functions 15
35. Thepopulation(inmillions) of acountryasafunctionof timet (years) isP(t ) = 30.2
0.1t
. Showthatthepopulation
doublesevery 10years. Showmoregenerally that for any positiveconstantsa andk, thefunctiong(t ) = a2
kt
doubles
after 1/k years.
solution Let P(t ) = 30· 2
0.1t
. Then
P(t +10) = 30· 2
0.1(t +10)
= 30· 2
0.1t +1
= 2(30· 2
0.1t
) = 2P(t ).
Hence, thepopulationdoublesinsizeevery10years. Inthemoregeneral case, let g(t ) = a2
kt
. Then
g
_
t +
1
k
_
= a2
k(t +1/k)
= a2
kt +1
= 2a2
kt
= 2g(t ).
Hence, thefunctiong doublesafter 1/k years.
Findall valuesof c suchthat f (x) =
x +1
x
2
+2cx +4
hasdomainR. Further Insights and Challenges
In Exercises 37–43, we deﬁne the ﬁrst differenceδf of a function f (x) by δf (x) = f (x +1) −f (x).
37. Showthat if f (x) = x
2
, thenδf (x) = 2x +1. Calculateδf for f (x) = x andf (x) = x
3
.
solution f (x) = x
2
: δf (x) = f (x +1) −f (x) = (x +1)
2
−x
2
= 2x +1
f (x) = x: δf (x) = x +1−x = 1
f (x) = x
3
: δf (x) = (x +1)
3
−x
3
= 3x
2
+3x +1
Showthat δ(10
x
) = 9· 10
x
and, moregenerally, that δ(b
x
) = (b −1)b
x
.
39. Showthat for anytwofunctionsf andg, δ(f +g) = δf +δg andδ(cf ) = cδ(f ), wherec isanyconstant.
solution δ(f +g) = (f (x +1) +g(x +1)) −(f (x) −g(x))
= (f (x +1) −f (x)) +(g(x +1) −g(x)) = δf (x) +δg(x)
δ(cf ) = cf (x +1) −cf (x) = c(f (x +1) −f (x)) = cδf (x).
SupposewecanﬁndafunctionP(x) suchthat δP = (x + 1)
k
andP(0) = 0. Provethat P(1) = 1
k
, P(2) =
1
k
+2
k
, and, moregenerally, for everywholenumber n,
P(n) = 1
k
+2
k
+· · · +n
k
41. First showthat
P(x) =
x(x +1)
2
satisﬁesδP = (x +1). ThenapplyExercise40toconcludethat
1+2+3+· · · +n =
n(n +1)
2
solution Let P(x) = x(x +1)/2. Then
δP(x) = P(x +1) −P(x) =
(x +1)(x +2)
2

x(x +1)
2
=
(x +1)(x +2−x)
2
= x +1.
Also, notethat P(0) = 0. Thus, byExercise40, withk = 1, it followsthat
P(n) =
n(n +1)
2
= 1+2+3+· · · +n.
Calculateδ(x
3
), δ(x
2
), andδ(x).Thenﬁndapolynomial P(x) of degree3suchthatδP = (x +1)
2
andP(0) = 0.
Concludethat P(n) = 1
2
+2
2
+· · · +n
2
.
43. ThisexercisecombinedwithExercise40showsthatforall wholenumbersk, thereexistsapolynomial P(x) satisfying
Eq. (1). ThesolutionrequirestheBinomial Theoremandproof byinduction(seeAppendixC).
(a) Showthat δ(x
k+1
) = (k +1) x
k
+· · · , wherethedotsindicatetermsinvolvingsmaller powersof x.
(b) Showbyinductionthat thereexistsapolynomial of degreek +1withleadingcoefﬁcient 1/(k +1):
P(x) =
1
k +1
x
k+1
+· · ·
suchthat δP = (x +1)
k
andP(0) = 0.
solution
(a) BytheBinomial Theorem:
δ(x
n+1
) = (x +1)
n+1
−x
n+1
=
_
x
n+1
+
_
n +1
1
_
x
n
+
_
n +1
2
_
x
n−1
+· · · +1
_
−x
n+1
=
_
n +1
1
_
x
n
+
_
n +1
2
_
x
n−1
+· · · +1
Thus,
δ(x
n+1
) = (n +1) x
n
+· · ·
wherethedotsindicatetermsinvolvingsmaller powersof x.
June 7, 2011 LTSV SSM Second Pass
16 C HA P T E R 1 PRECALCULUS REVIEW
(b) For k = 0, notethat P(x) = x satisﬁesδP = (x +1)
0
= 1andP(0) = 0.
Nowsupposethepolynomial
P(x) =
1
k
x
k
+p
k−1
x
k−1
+· · · +p
1
x
whichclearlysatisﬁesP(0) = 0alsosatisﬁesδP = (x +1)
k−1
. Wetrytoprovetheexistenceof
Q(x) =
1
k +1
x
k+1
+q
k
x
k
+· · · +q
1
x
suchthat δQ = (x +1)
k
. Observethat Q(0) = 0.
If δQ = (x +1)
k
andδP = (x +1)
k−1
, then
δQ = (x +1)
k
= (x +1)δP = xδP(x) +δP
Bythelinearityof δ (Exercise39), weﬁndδQ−δP = xδP or δ(Q−P) = xδP. Bydeﬁnition,
Q−P =
1
k +1
x
k+1
+
_
q
k

1
k
_
x
k
+· · · +(q
1
−p
1
)x,
so, bythelinearityof δ,
δ(Q−P) =
1
k +1
δ(x
k+1
) +
_
q
k

1
k
_
δ(x
k
) +· · · +(q
1
−p
1
) = x(x +1)
k−1
Bypart (a),
δ(x
k+1
) = (k +1)x
k
+L
k−1,k−1
x
k−1
+. . . +L
k−1,1
x +1
δ(x
k
) = kx
k−1
+L
k−2,k−2
x
k−2
+. . . +L
k−2,1
x +1
.
.
.
δ(x
2
) = 2x +1
wheretheL
i,j
arereal numbersfor eachi, j.
Toconstruct Q, wehavetogrouplikepowersof x onbothsidesof Eq. (43b). Thisyieldsthesystemof equations
1
k +1
_
(k +1)x
k
_
= x
k
1
k +1
L
k−1,k−1
x
k−1
+
_
q
k

1
k
_
kx
k−1
= (k −1)x
k−1
.
.
.
1
k +1
+
_
q
k

1
k
_
+(q
k−1
−p
k−1
) +· · · +(q
1
−p
1
) = 0.
Theﬁrst equationis identically true, andthesecondequationcanbesolvedimmediately for q
k
. Substitutingthevalue
of q
k
intothethirdequationof thesystem, wecanthensolvefor q
k−1
. Wecontinuethisprocessuntil wesubstitutethe
valuesof q
k
, q
k−1
, . . . q
2
intothelast equation, andthensolvefor q
1
.
1.4 Trigonometric Functions
Preliminary Questions
1. Howisit possiblefor twodifferent rotationstodeﬁnethesameangle?
solution Workingfromthesameinitial radius, two rotations that differ by awholenumber of full revolutions will
havethesameendingradius; consequently, thetworotationswill deﬁnethesameangleeventhoughthemeasuresof the
rotationswill bedifferent.
2. Givetwodifferent positiverotationsthat deﬁnetheangleπ/4.
solution Theangleπ/4is deﬁnedby any rotationof theform
π
4
+ 2πk wherek is aninteger. Thus, two different
positiverotationsthat deﬁnetheangleπ/4are
π
4
+2π(1) =

4
and
π
4
+2π(5) =
41π
4
.
June 7, 2011 LTSV SSM Second Pass
S E C T I ON 1.4 Trigonometric Functions 17
3. Giveanegativerotationthat deﬁnestheangleπ/3.
solution Theangleπ/3isdeﬁnedbyanyrotationof theform
π
3
+2πk wherek isaninteger. Thus, anegativerotation
that deﬁnestheangleπ/3is
π
3
+2π(−1) = −

3
.
4. Thedeﬁnitionof cosθ usingright trianglesapplieswhen(choosethecorrect answer):
(a) 0< θ <
π
2
(b) 0< θ < π (c) 0< θ < 2π
solution Thecorrect responseis(a): 0< θ <
π
2
.
5. What istheunit circledeﬁnitionof sinθ?
solution Let O denotethecenter of theunit circle, andlet P beapoint ontheunit circlesuchthat theradius OP
makesanangleθ withthepositivex-axis. Then, sinθ isthey-coordinateof thepoint P.
6. Howdoestheperiodicityof sinθ andcosθ followfromtheunit circledeﬁnition?
solution Let O denotethecenter of theunit circle, andlet P beapoint ontheunit circlesuchthat theradiusOP
makesanangleθ withthepositivex-axis. Then, cosθ andsinθ arethex- andy-coordinates, respectively, of thepoint
P. Theangleθ +2π isobtainedfromtheangleθ bymakingonefull revolutionaroundthecircle. Theangleθ +2π will
thereforehavetheradiusOP asitsterminal side. Thus
cos(θ +2π) = cosθ and sin(θ +2π) = sinθ.
Inother words, sinθ andcosθ areperiodicfunctions.
Exercises
1. Findtheanglebetween0and2π equivalent to13π/4.
solution Because13π/4> 2π, werepeatedlysubtract 2π until wearriveat aradianmeasurethat isbetween0and
2π. After onesubtraction, wehave13π/4−2π = 5π/4. Because0< 5π/4< 2π, 5π/4istheanglemeasurebetween
0and2π that isequivalent to13π/4.
Describeθ = π/6byanangleof negativeradianmeasure.
3. Convert fromradianstodegrees:
(a) 1 (b)
π
3
(c)
5
12
(d) −

4
solution
(a) 1
_
180

π
_
=
180

π
≈ 57.3

(b)
π
3
_
180

π
_
= 60

(c)
5
12
_
180

π
_
=
75

π
≈ 23.87

(d) −

4
_
180

π
_
= −135

Convert fromdegreestoradians:
(a) 1

(b) 30

(c) 25

(d) 120

5. Findthelengthsof thearcssubtendedbytheanglesθ andφ radiansinFigure20.
4
q = 0.9
f = 2
FIGURE 20 Circleof radius4.
solution s = rθ = 4(.9) = 3.6; s = rφ = 4(2) = 8
Calculatethevaluesof thesixstandardtrigonometricfunctionsfor theangleθ inFigure21.
7. Fill intheremainingvaluesof (cosθ, sinθ) for thepointsinFigure22.
p
2
0 (0, 0) p
5p
6
7p
6
11p
6
3p
4
5p
4
7p
4 4p
3
5p
3
3p
2
2p
3
( , )
p
6

2
3 1
2
( , )
2
3 1
2
p
3
( , )
p
4

2
2
2
2
FIGURE 22
June 7, 2011 LTSV SSM Second Pass
18 C HA P T E R 1 PRECALCULUS REVIEW
solution
θ
π
2

3

4

6
π

6
(cosθ, sinθ) (0, 1)
_
−1
2
,

3
2
_ _

2
2
,

2
2
_ _

3
2
,
1
2
_
(−1, 0)
_

3
2
,
−1
2
_
θ

4

3

2

3

4
11π
6
(cosθ, sinθ)
_

2
2
,

2
2
_ _
−1
2
,

3
2
_
(0, −1)
_
1
2
,

3
2
_ _√
2
2
,

2
2
_ _√
3
2
,
−1
2
_
Findthevaluesof thesixstandardtrigonometricfunctionsat θ = 11π/6.
In Exercises 9–14, use Figure 22 to ﬁnd all angles between 0and 2π satisfying the given condition.
9. cosθ =
1
2
solution θ =
π
3
,

3
tanθ = 1
11. tanθ = −1
solution θ =

4
,

4
cscθ = 2
13. sinx =

3
2
solution x =
π
3
,

3
sect = 2
15. Fill inthefollowingtableof values:
θ
π
6
π
4
π
3
π
2

3

4

6
tanθ
secθ
solution
θ
π
6
π
4
π
3
π
2

3

4

6
tanθ
1

3
1

3 und −

3 −1 −
1

3
secθ
2

3

2 2 und −2 −

2 −
2

3
Completethefollowingtableof signs:
θ sinθ cosθ tanθ cotθ secθ cscθ
0< θ <
π
2
+ +
π
2
< θ < π
π < θ <

2

2
< θ < 2π
17. Showthat if tanθ = c and0≤ θ < π/2, thencosθ = 1/
_
1+c
2
. Hint: Drawaright trianglewhoseoppositeand
adjacent sideshavelengthsc and1.
solution Because0≤ θ < π/2, wecanusethedeﬁnitionof thetrigonometric functionsintermsof right triangles.
tanθ is theratioof thelengthof thesideoppositetheangleθ tothelengthof theadjacent side. Withc =
c
1
, welabel
thelengthof theoppositesideasc andthelengthof theadjacent sideas1(seethediagrambelow). BythePythagorean
theorem, thelengthof thehypotenuseis
_
1+c
2
. Finally, weusethefactthatcosθ istheratioof thelengthof theadjacent
sidetothelengthof thehypotenusetoobtain
cosθ =
1
_
1+c
2
.
q
1+ c
2
c
1
Supposethat cosθ =
1
3
.
(a) Showthat if 0≤ θ < π/2, thensinθ = 2

2/3andtanθ = 2

2.
(b) Findsinθ andtanθ if 3π/2≤ θ < 2π.
June 7, 2011 LTSV SSM Second Pass
S E C T I ON 1.4 Trigonometric Functions 19
In Exercises 19–24, assume that 0≤ θ < π/2.
19. Findsinθ andtanθ if cosθ =
5
13
.
solution Consider thetrianglebelow. Thelengthsof thesideadjacent totheangleθ andthehypotenusehavebeen
labeledsothatcosθ =
5
13
. Thelengthof thesideoppositetheangleθ hasbeencalculatedusingthePythagoreantheorem:
_
13
2
−5
2
= 12. Fromthetriangle, weseethat
sinθ =
12
13
and tanθ =
12
5
.
θ
5
12 13
Findcosθ andtanθ if sinθ =
3
5
.
21. Findsinθ, secθ, andcotθ if tanθ =
2
7
.
solution If tanθ =
2
7
, thencotθ =
7
2
. For theremainingtrigonometric functions, consider thetrianglebelow. The
lengthsof thesidesoppositeandadjacenttotheangleθ havebeenlabeledsothattanθ =
2
7
. Thelengthof thehypotenuse
hasbeencalculatedusingthePythagoreantheorem:
_
2
2
+7
2
=

53. Fromthetriangle, weseethat
sinθ =
2

53
=
2

53
53
and secθ =

53
7
.
2
q
53
7
Findsinθ, cosθ, andsecθ if cotθ = 4.
23. Findcos2θ if sinθ =
1
5
.
solution Usingthedoubleangleformulacos2θ = cos
2
θ −sin
2
θ andthefundamental identitysin
2
θ +cos
2
θ = 1,
weﬁndthat cos2θ = 1−2sin
2
θ. Thus, cos2θ = 1−2(1/25) = 23/25.
Findsin2θ andcos2θ if tanθ =

2.
25. Findcosθ andtanθ if sinθ = 0.4andπ/2≤ θ < π.
solution Wecandeterminethe“magnitude” of cosθ andtanθ usingthetriangleshownbelow. Thelengths of the
sideoppositetheangleθ andthehypotenusehavebeenlabeledsothat sinθ = 0.4=
2
5
. Thelengthof thesideadjacent
totheangleθ wascalculatedusingthePythagoreantheorem:
_
5
2
−2
2
=

21. Fromthetriangle, weseethat
|cosθ| =

21
5
and |tanθ| =
2

21
=
2

21
21
.
Becauseπ/2≤ θ < π, bothcosθ andtanθ arenegative; consequently,
cosθ = −

21
5
and tanθ = −
2

21
21
.
2
5
q
21
Findcosθ andsinθ if tanθ = 4andπ ≤ θ < 3π/2.
27. Findcosθ if cotθ =
4
3
andsinθ < 0.
solution Wecandeterminethe“magnitude”of cosθ usingthetriangleshownbelow. Thelengthsof thesidesopposite
andadjacent totheangleθ havebeenlabeledsothat cotθ =
4
3
. Thelengthof thehypotenusewascalculatedusingthe
Pythagoreantheorem:
_
3
2
+4
2
= 5. Fromthetriangle, weseethat
|cosθ| =
4
5
.
Becausecotθ =
4
3
> 0andsinθ < 0, theangleθ mustbeinthethirdquadrant; consequently, cosθ will benegativeand
cosθ = −
4
5
.
June 7, 2011 LTSV SSM Second Pass
20 C HA P T E R 1 PRECALCULUS REVIEW
4
3
5

θ
Findtanθ if secθ =

5andsinθ < 0.
29. Findthevaluesof sinθ, cosθ, andtanθ for theanglescorrespondingtotheeight pointsinFigure23(A) and(B).
(0.3965, 0.918)
(A) (B)
(0.3965, 0.918)
FIGURE 23
solution Let’sstart withthefour pointsinFigure23(A).
• Thepoint intheﬁrst quadrant hascoordinates(0.3965, 0.918). Therefore,
sinθ = 0.918, cosθ = 0.3965, and tanθ =
0.918
0.3965
= 2.3153.
• Thecoordinatesof thepoint inthesecondquadrant are(−0.918, 0.3965). Therefore,
sinθ = 0.3965, cosθ = −0.918, and tanθ =
0.3965
−0.918
= −0.4319.
• Becausethepoint inthethirdquadrant issymmetrictothepoint intheﬁrst quadrant withrespect totheorigin, its
coordinatesare(−0.3965, −0.918). Therefore,
sinθ = −0.918, cosθ = −0.3965, and tanθ =
−0.918
−0.3965
= 2.3153.
• Becausethepoint inthefourthquadrant issymmetrictothepoint inthesecondquadrant withrespect totheorigin,
itscoordinatesare(0.918, −0.3965). Therefore,
sinθ = −0.3965, cosθ = 0.918, and tanθ =
−0.3965
0.918
= −0.4319.
Nowconsider thefour pointsinFigure23(B).
• Thepoint intheﬁrst quadrant hascoordinates(0.3965, 0.918). Therefore,
sinθ = 0.918, cosθ = 0.3965, and tanθ =
0.918
0.3965
= 2.3153.
• Thepoint inthesecondquadrant isareﬂectionthroughthey-axisof thepoint intheﬁrst quadrant. Itscoordinates
aretherefore(−0.3965, 0.918) and
sinθ = 0.918, cosθ = −0.3965, and tanθ =
0.918
0.3965
= −2.3153.
• Becausethepoint inthethirdquadrant issymmetrictothepoint intheﬁrst quadrant withrespect totheorigin, its
coordinatesare(−0.3965, −0.918). Therefore,
sinθ = −0.918, cosθ = −0.3965, and tanθ =
−0.918
−0.3965
= 2.3153.
• Becausethepoint inthefourthquadrant issymmetrictothepoint inthesecondquadrant withrespect totheorigin,
itscoordinatesare(0.3965, −0.918). Therefore,
sinθ = −0.918, cosθ = 0.3965, and tanθ =
−0.918
0.3965
= −2.3153.
June 7, 2011 LTSV SSM Second Pass
S E C T I ON 1.4 Trigonometric Functions 21
Refer toFigure24(A). Expressthefunctionssinθ, tanθ, andcscθ intermsof c.
31. Refer toFigure24(B). Computecosψ, sinψ, cotψ, andcscψ.
c
1 1
0.3
(B) (A)
θ ψ
FIGURE 24
solution By thePythagoreantheorem, thelengthof thesideoppositetheangleψ inFigure24(B) is
_
1−0.3
2
=

0.91. Consequently,
cosψ =
0.3
1
= 0.3, sinψ =

0.91
1
=

0.91, cotψ =
0.3

0.91
and cscψ =
1

0.91
.
Expresscos
_
θ +
π
2
_
andsin
_
θ +
π
2
_
intermsof cosθ andsinθ. Hint: Findtherelationbetweenthecoordinates
(a, b) and(c, d) inFigure25.
33. Usetheadditionformulatocomputecos
_
π
3
+
π
4
_
exactly.
solution
cos
_
π
3
+
π
4
_
= cos
π
3
cos
π
4
−sin
π
3
sin
π
4
=
1
2
·

2
2

3
2
·

2
2
=

2−

6
4
.
Usetheadditionformulatocomputesin
_
π
3

π
4
_
exactly.
In Exercises 35–38, sketch the graph over [0, 2π].
35. 2sin4θ
solution
−2
−1
2
1
y
x
6 5 4 3 2 1
cos
_
2
_
θ −
π
2
__
37. cos
_
2θ −
π
2
_
solution
−1
−0.5
1
0.5
y
x
6 5 4 3 2 1
sin
_
2
_
θ −
π
2
_

_
+2
39. Howmany pointslieontheintersectionof thehorizontal liney = c andthegraphof y = sinx for 0 ≤ x < 2π?
Hint: Theanswer dependsonc.
solution Recall that for anyx, −1≤ sinx ≤ 1. Thus, if |c| > 1, thehorizontal liney = c andthegraphof y = sinx
never intersect. If c = +1, theny = c andy = sinx intersect at thepeak of thesinecurve; that is, they intersect at
x =
π
2
. Ontheother hand, if c = −1, theny = c andy = sinx intersect at thebottomof thesinecurve; that is, they
intersect at x =

2
. Finally, if |c| < 1, thegraphsof y = c andy = sinx intersect twice.
Howmanypointslieontheintersectionof thehorizontal liney = c andthegraphof y = tanx for 0≤ x < 2π?
In Exercises 41–44, solve for 0≤ θ < 2π (see Example 4).
41. sin2θ +sin3θ = 0
solution sinα = −sinβ whenα = −β + 2πk or α = π + β + 2πk. Substitutingα = 2θ andβ = 3θ, wehave
either2θ = −3θ +2πk or2θ = π +3θ +2πk. Solvingeachof theseequationsforθ yieldsθ =
2
5
πk orθ = −π −2πk.
Thesolutionsontheinterval 0≤ θ < 2π arethen
θ = 0,

5
,

5
, π,

5
,

5
.
sinθ = sin2θ
June 7, 2011 LTSV SSM Second Pass
22 C HA P T E R 1 PRECALCULUS REVIEW
43. cos4θ +cos2θ = 0
solution cosα = −cosβ whenα +β = π +2πk or α = β +π +2πk. Substitutingα = 4θ andβ = 2θ, wehave
either 6θ = π +2πk or 4θ = 2θ +π +2πk. Solvingeachof theseequationsfor θ yieldsθ =
π
6
+
π
3
k or θ =
π
2
+πk.
Thesolutionsontheinterval 0≤ θ < 2π arethen
θ =
π
6
,
π
2
,

6
,

6
,

2
,
11π
6
.
sinθ = cos2θ
In Exercises 45–54, derive the identity using the identities listed in this section.
45. cos2θ = 2cos
2
θ −1
solution Startingfromthedoubleangleformulafor cosine, cos
2
θ =
1
2
(1+cos2θ), wesolvefor cos2θ. Thisgives
2cos
2
θ = 1+cos2θ andthencos2θ = 2cos
2
θ −1.
cos
2
θ
2
=
1+cosθ
2
47. sin
θ
2
=
_
1−cosθ
2
solution Substitutex = θ/2intothedoubleangleformulafor sine, sin
2
x =
1
2
(1−cos2x) toobtainsin
2
_
θ
2
_
=
1−cosθ
2
. Takingthesquareroot of bothsidesyieldssin
_
θ
2
_
=
_
1−cosθ
2
.
sin(θ +π) = −sinθ
49. cos(θ +π) = −cosθ
solution Fromtheadditionformulafor thecosinefunction, wehave
cos(θ +π) = cosθ cosπ −sinθ sinπ = cosθ(−1) = −cosθ
tanx = cot
_
π
2
−x
_ 51. tan(π −θ) = −tanθ
solution UsingExercises48and49,
tan(π −θ) =
sin(π −θ)
cos(π −θ)
=
sin(π +(−θ))
cos(π +(−θ))
=
−sin(−θ)
−cos(−θ)
=
sinθ
−cosθ
= −tanθ.
Thesecondtolast equalityoccursbecausesinx isanoddfunctionandcosx isanevenfunction.
tan2x =
2tanx
1−tan
2
x
53. tanx =
sin2x
1+cos2x
solution Usingtheadditionformulafor thesinefunction, weﬁnd
sin2x = sin(x +x) = sinx cosx +cosx sinx = 2sinx cosx.
ByExercise45, weknowthat cos2x = 2cos
2
x −1. Therefore,
sin2x
1+cos2x
=
2sinx cosx
1+2cos
2
x −1
=
2sinx cosx
2cos
2
x
=
sinx
cosx
= tanx.
sin
2
x cos
2
x =
1−cos4x
8
55. UseExercises48and49toshowthat tanθ andcotθ areperiodicwithperiodπ.
solution ByExercises48and49,
tan(θ +π) =
sin(θ +π)
cos(θ +π)
=
−sinθ
−cosθ
= tanθ,
and
cot(θ +π) =
cos(θ +π)
sin(θ +π)
=
−cosθ
−sinθ
= cotθ.
Thus, bothtanθ andcotθ areperiodicwithperiodπ.
Usetheidentityof Exercise45toshowthat cos
π
8
isequal to
_
1
2
+

2
4
.
57. UsetheLawof CosinestoﬁndthedistancefromP toQinFigure26.
P
Q
8
10
7π/9
FIGURE 26
solution BytheLawof Cosines, thedistancefromP toQis
_
10
2
+8
2
−2(10)(8) cos

9
= 16.928.
June 7, 2011 LTSV SSM Second Pass
S E C T I ON 1.5 Technology: Calculators and Computers 23
Further Insights and Challenges
UseFigure27toderivetheLawof CosinesfromthePythagoreanTheorem.
59. Usetheadditionformulatoprove
cos3θ = 4cos
3
θ −3cosθ
solution
cos3θ = cos(2θ +θ) = cos2θ cosθ −sin2θ sinθ = (2cos
2
θ −1) cosθ −(2sinθ cosθ) sinθ
= cosθ(2cos
2
θ −1−2sin
2
θ) = cosθ(2cos
2
θ −1−2(1−cos
2
θ))
= cosθ(2cos
2
θ −1−2+2cos
2
θ) = 4cos
3
θ −3cosθ
Usetheadditionformulasfor sineandcosinetoprove
tan(a +b) =
tana +tanb
1−tana tanb
cot(a −b) =
cota cotb +1
cotb −cota
61. Let θ betheanglebetweentheliney = mx +b andthex-axis[Figure28(A)]. Provethat m = tanθ.
y = mx + b
q
x
r
s
(A)
y
q
x
(B)
y
L
2
L
1
FIGURE 28
solution UsingthedistanceslabeledinFigure28(A), weseethat theslopeof thelineisgivenbytheratior/s. The
tangent of theangleθ isgivenbythesameratio. Therefore, m = tanθ.
LetL
1
andL
2
bethelinesof slopem
1
andm
2
[Figure28(B)]. Showthattheangleθ betweenL
1
andL
2
satisﬁes
cotθ =
m
2
m
1
+1
m
2
−m
1
.
63. Perpendicular Lines UseExercise62toprovethat twolineswithnonzeroslopesm
1
andm
2
areperpendicular if
andonlyif m
2
= −1/m
1
.
solution If linesareperpendicular, thentheanglebetweenthemisθ = π/2⇒
cot(π/2) =
1+m
1
m
2
m
1
−m
2
0=
1+m
1
m
2
m
1
−m
2
⇒ m
1
m
2
= −1⇒ m
1
= −
1
m
2
Applythedouble-angleformulatoprove:
(a) cos
π
8
=
1
2
_
2+

2
(b) cos
π
16
=
1
2
_
2+
_
2+

2
Guessthevaluesof cos
π
32
andof cos
π
2
n
for all n.
1.5 Technology: Calculators and Computers
Preliminary Questions
1. Isthereadeﬁnitewayof choosingtheoptimal viewingrectangle, or isit best toexperiment until youﬁndaviewing
rectangleappropriatetotheproblemat hand?
solution It isbest toexperiment withthewindowsizeuntil oneisfoundthat isappropriatefor theproblemat hand.
2. Describethecalculator screenproducedwhenthefunctiony = 3+x
2
isplottedwithviewingrectangle:
(a) [−1, 1] ×[0, 2] (b) [0, 1] ×[0, 4]
solution
(a) Usingtheviewingrectangle[−1, 1] by[0, 2], thescreenwill displaynothingastheminimumvalueof y = 3+x
2
isy = 3.
(b) Usingtheviewingrectangle[0, 1] by [0, 4], thescreenwill display theportionof theparabolabetweenthepoints
(0, 3) and(1, 4).
3. AccordingtotheevidenceinExample4, it appearsthat f (n) = (1+1/n)
n
never takesonavaluegreater than3for
n > 0. Doesthisevidenceprove that f (n) ≤ 3for n > 0?
solution No, thisevidencedoesnot constituteaproof that f (n) ≤ 3for n ≥ 0.
4. Howcanagraphingcalculator beusedtoﬁndtheminimumvalueof afunction?
solution Experiment with theviewing window to zoomin on thelowest point on thegraph of thefunction. The
y-coordinateof thelowest point onthegraphistheminimumvalueof thefunction.
June 7, 2011 LTSV SSM Second Pass
24 C HA P T E R 1 PRECALCULUS REVIEW
Exercises
The exercises in this section should be done using a graphing calculator or computer algebra system.
1. Plot f (x) = 2x
4
+3x
3
−14x
2
−9x +18intheappropriateviewingrectanglesanddetermineitsroots.
solution Usingaviewingrectangleof [−4, 3] by[−20, 20], weobtaintheplot below.
−10
−20
20
10
y
x
−4 −2 −3 −1 1 2 3
Now, therootsof f (x) arethex-interceptsof thegraphof y = f (x). Fromtheplot, wecanidentifythex-interceptsas
−3, −1.5, 1, and2. Therootsof f (x) arethereforex = −3, x = −1.5, x = 1, andx = 2.
Howmanysolutionsdoesx
3
−4x +8= 0have?
3. Howmanypositive solutionsdoesx
3
−12x +8= 0have?
solution Thegraphof y = x
3
− 12x + 8shownbelowhastwox-interceptstotheright of theorigin; thereforethe
equationx
3
−12x +8= 0hastwopositivesolutions.
−20
−40
−60
60
40
20
y
x
−4 −2 4 2
Doescosx +x = 0haveasolution?A positivesolution?
5. Findall thesolutionsof sinx =

x for x > 0.
solution Solutionstotheequationsinx =

x correspondtopointsof intersectionbetweenthegraphsof y = sinx
andy =

x. Thetwographsareshownbelow; theonlypointof intersectionisatx = 0. Therefore, therearenosolutions
of sinx =

x for x > 0.
x
1
2
5 4 3 2 1
y
−1
Howmanysolutionsdoescosx = x
2
have?
7. Let f (x) = (x − 100)
2
+ 1000. What will thedisplay showif yougraphf (x) intheviewingrectangle[−10, 10]
by[−10, 10]? Findanappropriateviewingrectangle.
solution Because(x −100)
2
≥ 0for all x, it followsthat f (x) = (x −100)
2
+1000≥ 1000for all x. Thus, using
aviewingrectangleof [−10, 10] by [−10, 10] will display nothing. Theminimumvalueof thefunctionoccurs when
x = 100, soanappropriateviewingrectanglewouldbe[50, 150] by[1000, 2000].
Plot f (x) =
8x +1
8x −4
inanappropriateviewingrectangle. What arethevertical andhorizontal asymptotes?
9. Plotthegraphof f (x) = x/(4−x) inaviewingrectanglethatclearlydisplaysthevertical andhorizontal asymptotes.
solution Fromthegraphof y =
x
4−x
shownbelow, weseethat thevertical asymptoteisx = 4andthehorizontal
asymptoteisy = −1.
−2
2
y
x
−8 −4
4 8 12 16
Illustratelocal linearityfor f (x) = x
2
byzoominginonthegraphat x = 0.5(seeExample6).
11. Plotf (x) = cos(x
2
) sinx for 0≤ x ≤ 2π. Thenillustratelocal linearityatx = 3.8bychoosingappropriateviewing
rectangles.
solution The following three graphs display f (x) = cos(x
2
) sinx over the intervals [0, 2π], [3.5, 4.1] and
[3.75, 3.85]. Theﬁnal graphlookslikeastraight line.
June 7, 2011 LTSV SSM Second Pass
S E C T I ON 1.5 Technology: Calculators and Computers 25
x
1
−1
1 2 3 4 5 6
y
x
1
−1
3.5 3.6 3.7 3.8 3.9 4
y
x
−0.2
0.4
0.2
3.76 3.8 3.78 3.82 3.84
y
If P
0
dollarsaredepositedinabankaccountpaying5%interestcompoundedmonthly, thentheaccounthasvalue
P
0
_
1+
0.05
12
_
N
after N months. Find, tothenearestinteger N, thenumber of monthsafter whichtheaccountvalue
doubles.
In Exercises 13–18, investigate the behavior of the function as n or x grows large by making a table of function values
and plotting a graph (see Example 4). Describe the behavior in words.
13. f (n) = n
1/n
solution Thetableandgraphsbelowsuggest that asn getslarge, n
1/n
approaches1.
n n
1/n
10 1.258925412
10
2
1.047128548
10
3
1.006931669
10
4
1.000921458
10
5
1.000115136
10
6
1.000013816
x
y
1
0 2 4 6 8 10
x
y
1
0 200 400 600 800 1000
f (n) =
4n +1
6n −5
15. f (n) =
_
1+
1
n
_
n
2
solution Thetableandgraphsbelowsuggest that asn getslarge, f (n) tendstoward∞.
n
_
1+
1
n
_
n
2
10 13780.61234
10
2
1.635828711×10
43
10
3
1.195306603×10
434
10
4
5.341783312×10
4342
10
5
1.702333054×10
43429
10
6
1.839738749×10
434294
x
y
10,000
0 2 4 6 8 10
x
y
1 × 10
43
0 20 40 60 80 100
f (x) =
_
x +6
x −4
_
x
17. f (x) =
_
x tan
1
x
_
x
solution Thetableandgraphsbelowsuggest that asx getslarge, f (x) approaches1.
x
_
x tan
1
x
_
x
10 1.033975759
10
2
1.003338973
10
3
1.000333389
10
4
1.000033334
10
5
1.000003333
10
6
1.000000333
June 7, 2011 LTSV SSM Second Pass
26 C HA P T E R 1 PRECALCULUS REVIEW
x
y
1
1.1
1.2
1.3
1.4
1.5
5 10 15 20
x
20 40 60 80 100
y
1
1.1
1.2
1.3
1.4
1.5
f (x) =
_
x tan
1
x
_
x
2
19. Thegraphof f (θ) = Acosθ + B sinθ isasinusoidal wavefor any constantsA andB. Conﬁrmthisfor (A, B) =
(1, 1), (1, 2), and(3, 4) byplottingf (θ).
solution Thegraphsof f (θ) = cosθ +sinθ, f (θ) = cosθ +2sinθ andf (θ) = 3cosθ +4sinθ areshownbelow.
y
x
−2 2
(A, B) = (1, 1)
4 6 8
1
−1
y
x
−2 2
(A, B) = (1, 2)
4 6 8
2
1
−2
−1
y
x
−2 2
(A, B) = (3, 4)
4 6 8
4
2
−4
−2
Find themaximumvalueof f (θ) for thegraphs produced in Exercise19. Can you guess theformulafor the
maximumvalueintermsof A andB?
21. Findtheintervalsonwhichf (x) = x(x +2)(x −3) ispositivebyplottingagraph.
solution Thefunctionf (x) = x(x + 2)(x − 3) is positivewhenthegraphof y = x(x + 2)(x − 3) lies abovethe
x-axis. Thegraphof y = x(x + 2)(x − 3) isshownbelow. Clearly, thegraphliesabovethex-axisandthefunctionis
positivefor x ∈ (−2, 0) ∪ (3, ∞).
−20
−40
20
y
x
−4 −2 2 4
Findtheset of solutionstotheinequality(x
2
−4)(x
2
−1) < 0byplottingagraph.
Further Insights and Challenges
23. Let f
1
(x) = x and deﬁne a sequence of functions by f
n+1
(x) =
1
2
(f
n
(x) + x/f
n
(x)). For example,
f
2
(x) =
1
2
(x +1). Useacomputer algebrasystemtocomputef
n
(x) for n = 3, 4, 5andplotf
n
(x) together with

x for
x ≥ 0. What doyounotice?
solution Withf
1
(x) = x andf
2
(x) =
1
2
(x +1), wecalculate
f
3
(x) =
1
2
_
1
2
(x +1) +
x
1
2
(x +1)
_
=
x
2
+6x +1
4(x +1)
f
4
(x) =
1
2

x
2
+6x +1
4(x +1)
+
x
x
2
+6x+1
4(x+1)

=
x
4
+28x
3
+70x
2
+28x +1
8(1+x)(1+6x +x
2
)
and
f
5
(x) =
1+120x +1820x
2
+8008x
3
+12870x
4
+8008x
5
+1820x
6
+120x
7
+x
8
16(1+x)(1+6x +x
2
)(1+28x +70x
2
+28x
3
+x
4
)
.
A plot of f
1
(x), f
2
(x), f
3
(x), f
4
(x), f
5
(x) and

x isshownbelow, withthegraphof

x shownasadashedcurve. It
seemsasif thef
n
areasymptoticto

x.
y
x
4
8
12
40 20 60 80 100
Set P
0
(x) = 1 and P
1
(x) = x. The Chebyshev polynomials (useful in approximation theory) are deﬁned
inductivelybytheformulaP
n+1
(x) = 2xP
n
(x) −P
n−1
(x).
(a) Showthat P
2
(x) = 2x
2
−1.
(b) ComputeP
n
(x) for 3≤ n ≤ 6usingacomputer algebrasystemor byhand andplot P
n
(x) over [−1 1]
June 7, 2011 LTSV SSM Second Pass
Chapter Review Exercises 27
CHAPTER REVIEW EXERCISES
1. Express(4, 10) asaset {x : |x −a| < c} for suitablea andc.
solution Thecenter of theinterval (4, 10) is
4+10
2
= 7andtheradiusis
10−4
2
= 3. Therefore, theinterval (4, 10) is
equivalent totheset {x : |x −7| < 3}.
Expressasaninterval:
(a) {x : |x −5| < 4} (b) {x : |5x +3| ≤ 2}
3. Express{x : 2≤ |x −1| ≤ 6} asaunionof twointervals.
solution Theset {x : 2≤ |x − 1| ≤ 6} consistsof thosenumbersthat areat least 2but at most 6unitsfrom1. The
numbers larger than1that satisfy theseconditions are3 ≤ x ≤ 7, whilethenumbers smaller than1that satisfy these
conditionsare−5≤ x ≤ −1. Therefore{x : 2≤ |x −1| ≤ 6} = [−5, −1] ∪ [3, 7].
Giveanexampleof numbersx, y suchthat |x| +|y| = x −y.
5. Describethepairsof numbersx, y suchthat |x +y| = x −y.
solution First consider thecasewhenx +y ≥ 0. Then|x +y| = x +y andweobtaintheequationx +y = x −y.
Thesolutionof thisequationisy = 0. Thus, thepairs(x, 0) withx ≥ 0satisfy|x +y| = x −y. Next, consider thecase
whenx +y < 0. Then|x +y| = −(x +y) = −x −y andweobtaintheequation−x −y = x −y. Thesolutionof this
equationisx = 0. Thus, thepairs(0, y) withy < 0alsosatisfy|x +y| = x −y.
Sketchthegraphof y = f (x +2) −1, wheref (x) = x
2
for −2≤ x ≤ 2.
In Exercises 7–10, let f (x) be the function shown in Figure 1.
1 2 3 4
1
2
0
3
x
y
FIGURE 1
7. Sketchthegraphsof y = f (x) +2andy = f (x +2).
solution Thegraphof y = f (x) + 2isobtainedby shiftingthegraphof y = f (x) up2units(seethegraphbelow
at theleft). Thegraphof y = f (x + 2) isobtainedby shiftingthegraphof y = f (x) totheleft 2units(seethegraph
belowat theright).
x
yy
x
1
2
3
4
5
1 2 3 4
f (x) + 2
−1 −2
x
yy
x
1
2
3
4
5
1 2 3 4
f (x + 2)
−1 −2
Sketchthegraphsof y =
1
2
f (x) andy = f
_
1
2
x
_
.
9. Continuethegraphof f (x) totheinterval [−4, 4] asanevenfunction.
solution Tocontinuethegraphof f (x) totheinterval [−4, 4] asanevenfunction, reﬂect thegraphof f (x) across
they-axis(seethegraphbelow).
−1 −4 −2 −3
x
1 2 3 4
y
1
2
3
Continuethegraphof f (x) totheinterval [−4, 4] asanoddfunction.
In Exercises 11–14, ﬁnd the domain and range of the function.
11. f (x) =

x +1
solution Thedomainof thefunctionf (x) =

x +1is{x : x ≥ −1} andtherangeis{y : y ≥ 0}.
f (x) =
4
x
4
+1
13. f (x) =
2
3−x
solution Thedomainof thefunctionf (x) =
2
3−x
is{x : x = 3} andtherangeis{y : y = 0}.
f (x) =
_
x
2
−x +5
June 7, 2011 LTSV SSM Second Pass
28 C HA P T E R 1 PRECALCULUS REVIEW
15. Determinewhether thefunctionisincreasing, decreasing, or neither:
(a) f (x) = 3
−x
(b) f (x) =
1
x
2
+1
(c) g(t ) = t
2
+t (d) g(t ) = t
3
+t
solution
(a) Thefunctionf (x) = 3
−x
canberewrittenasf (x) = (
1
3
)
x
. Thisisanexponential functionwithabaselessthan1;
therefore, thisisadecreasingfunction.
(b) Fromthegraphof y = 1/(x
2
+1) shownbelow, weseethatthisfunctionisneither increasingnor decreasingfor all
x (thoughit isincreasingfor x < 0anddecreasingfor x > 0).
x
−3 −2 −1 1 2 3
y
0.2
0.4
0.6
0.8
1
(c) Thegraphof y = t
2
+t isanupwardopeningparabola; therefore, thisfunctionisneither increasingnor decreasing
for all t . By completingthesquareweﬁndy = (t +
1
2
)
2

1
4
. Thevertex of this parabolais thenat t = −
1
2
, so the
functionisdecreasingfor t < −
1
2
andincreasingfor t > −
1
2
.
(d) Fromthegraphof y = t
3
+t shownbelow, weseethat thisisanincreasingfunction.
−20
20
y
x
−1 1 2 3 −2 −3
Determinewhether thefunctioniseven, odd, or neither:
(a) f (x) = x
4
−3x
2
(b) g(x) = sin(x +1)
(c) f (x) = 2
−x
2
In Exercises 17–22, ﬁnd the equation of the line.
17. Linepassingthrough(−1, 4) and(2, 6)
solution Theslopeof thelinepassingthrough(−1, 4) and(2, 6) is
m =
6−4
2−(−1)
=
2
3
.
Theequationof thelinepassingthrough(−1, 4) and(2, 6) isthereforey −4=
2
3
(x +1) or 2x −3y = −14.
Linepassingthrough(−1, 4) and(−1, 6)
19. Lineof slope6through(9, 1)
solution Usingthepoint-slopeformfor theequationof aline, theequationof thelineof slope6andpassingthrough
(9, 1) isy −1= 6(x −9) or 6x −y = 53.
Lineof slope−
3
2
through(4, −12)
21. Linethrough(2, 3) parallel toy = 4−x
solution Theequation y = 4− x is in slope-intercept form; it follows that theslopeof this lineis −1. Any line
parallel toy = 4− x will havethesameslope, sowearelookingfor theequationof thelineof slope−1andpassing
through(2, 3). Theequationof thislineisy −3= −(x −2) or x +y = 5.
Horizontal linethrough(−3, 5)
23. Doesthefollowingtableof marketdatasuggestalinear relationshipbetweenpriceandnumber of homessoldduring
aone-year period? Explain.
Price(thousandsof \$) 180 195 220 240
No. of homessold 127 118 103 91
solution Examinetheslopebetweenconsecutivedatapoints. Theﬁrst pair of datapointsyieldsaslopeof
118−127
195−180
= −
9
15
= −
3
5
,
whilethesecondpair of datapointsyieldsaslopeof
103−118
220−195
= −
15
25
= −
3
5
June 7, 2011 LTSV SSM Second Pass
Chapter Review Exercises 29
andthelast pair of datapointsyieldsaslopeof
91−103
240−220
= −
12
20
= −
3
5
.
Becauseall threeslopes areequal, thedatadoes suggest alinear relationshipbetweenpriceandthenumber of homes
sold.
Doesthefollowingtableof revenuedatafor acomputer manufacturer suggest alinear relationbetweenrevenue
andtime? Explain.
Year 2001 2005 2007 2010
Revenue(billionsof \$) 13 18 15 11
25. Findtherootsof f (x) = x
4
−4x
2
andsketchitsgraph. Onwhichintervalsisf (x) decreasing?
solution Theroots of f (x) = x
4
− 4x
2
areobtainedby solvingtheequationx
4
− 4x
2
= x
2
(x − 2)(x + 2) = 0,
whichyieldsx = −2, x = 0andx = 2. Thegraphof y = f (x) isshownbelow. Fromthisgraphweseethat f (x) is
decreasingfor x lessthanapproximately−1.4andfor x between0andapproximately1.4.
10
20
y
x
−1 1
2 3 −2 −3
Let h(z) = 2z
2
+12z +3. Completethesquareandﬁndtheminimumvalueof h(z).
27. Let f (x) bethesquareof thedistancefromthepoint (2, 1) toapoint (x, 3x +2) ontheliney = 3x +2. Showthat
f (x) isaquadraticfunction, andﬁnditsminimumvaluebycompletingthesquare.
solution Letf (x) denotethesquareof thedistancefromthepoint(2, 1) toapoint(x, 3x +2) ontheliney = 3x +2.
Then
f (x) = (x −2)
2
+(3x +2−1)
2
= x
2
−4x +4+9x
2
+6x +1= 10x
2
+2x +5,
whichisaquadraticfunction. Completingthesquare, weﬁnd
f (x) = 10
_
x
2
+
1
5
x +
1
100
_
+5−
1
10
= 10
_
x +
1
10
_
2
+
49
10
.
Because(x +
1
10
)
2
≥ 0for all x, it followsthat f (x) ≥
49
10
for all x. Hence, theminimumvalueof f (x) is
49
10
.
Provethat x
2
+3x +3≥ 0for all x.
In Exercises 29–34, sketch the graph by hand.
29. y = t
4
solution
x
−1 −0.5 1 0.5
y
0.2
0.4
0.6
0.8
1
y = t
5 31. y = sin
θ
2
solution
y
x
−5 5 10
0.5
1
−0.5
−1
y = 10
−x
33. y = x
1/3
solution
x
−1 −2 −3 −4 1 2 3 4
y
1
2
−1
−2
y =
1
x
2
June 7, 2011 LTSV SSM Second Pass
30 C HA P T E R 1 PRECALCULUS REVIEW
35. Showthatthegraphof y = f
_
1
3
x −b
_
isobtainedbyshiftingthegraphof y = f
_
1
3
x
_
totheright3b units. Usethis
observationtosketchthegraphof y =
¸
¸
1
3
x −4
¸
¸
.
solution Let g(x) = f (
1
3
x). Then
g(x −3b) = f
_
1
3
(x −3b)
_
= f
_
1
3
x −b
_
.
Thus, thegraphof y = f (
1
3
x −b) isobtainedbyshiftingthegraphof y = f (
1
3
x) totheright 3b units.
Thegraphof y = |
1
3
x −4| isthegraphof y = |
1
3
x| shiftedright 12units(seethegraphbelow).
y
x
1
2
3
4
0 5 10 15 20
Leth(x) = cosx andg(x) = x
−1
. Computethecompositefunctionsh(g(x)) andg(h(x)), andﬁndtheirdomains.
37. Findfunctionsf andg suchthat thefunction
f (g(t )) = (12t +9)
4
solution Onepossiblechoiceisf (t ) = t
4
andg(t ) = 12t +9. Then
f (g(t )) = f (12t +9) = (12t +9)
4
asdesired.
Sketchthepoints ontheunit circlecorrespondingto thefollowingthreeangles, andﬁndthevalues of thesix
standardtrigonometricfunctionsat eachangle:
(a)

3
(b)

4
(c)

6
39. What istheperiodof thefunctiong(θ) = sin2θ +sin
θ
2
?
solution Thefunctionsin2θ hasaperiodof π, andthefunctionsin(θ/2) hasaperiodof 4π. Because4π isamultiple
of π, theperiodof thefunctiong(θ) = sin2θ +sinθ/2is4π.
Assumethat sinθ =
4
5
, whereπ/2< θ < π. Find:
(a) tanθ (b) sin2θ (c) csc
θ
2
41. Giveanexampleof valuesa, b suchthat
(a) cos(a +b) = cosa +cosb (b) cos
a
2
=
cosa
2
solution
(a) Takea = b = π/2. Thencos(a +b) = cosπ = −1but
cosa +cosb = cos
π
2
+cos
π
2
= 0+0= 0.
(b) Takea = π. Then
cos
_
a
2
_
= cos
_
π
2
_
= 0
but
cosa
2
=
cosπ
2
=
−1
2
= −
1
2
.
Let f (x) = cosx. Sketchthegraphof y = 2f
_
1
3
x −
π
4
_
for 0≤ x ≤ 6π.
43. Solvesin2x +cosx = 0for 0≤ x < 2π.
solution Using thedoubleangleformulafor thesinefunction, werewritetheequation as 2sinx cosx + cosx =
cosx(2sinx + 1) = 0. Thus, either cosx = 0 or sinx = −1/2. Fromhereweseethat thesolutions arex = π/2,
x = 7π/6, x = 3π/2andx = 11π/6.
Howdoesh(n) = n
2
/2
n
behavefor largewhole-number valuesof n? Doesh(n) tendtoinﬁnity?
45. Useagraphingcalculator todeterminewhether theequationcosx = 5x
2
−8x
4
hasanysolutions.
solution Thegraphsof y = cosx andy = 5x
2
− 8x
4
areshownbelow. Becausethegraphsdonot intersect, there
arenosolutionstotheequationcosx = 5x
2
−8x
4
.
x
−1 1
y
y = cos x
y = 5x
2
− 8x
4
1
−1
Usingagraphingcalculator, ﬁndthenumber of real rootsandestimatethelargest root totwodecimal places:
(a) f (x) = 1.8x
4
−x
5
−x
(b) g(x) = 1.7x
4
−x
5
−x
June 7, 2011 LTSV SSM Second Pass
2 LIMITS
2.1 Limits, Rates of Change, and Tangent Lines
Preliminary Questions
1. Averagevelocityisequal totheslopeof asecant linethroughtwopointsonagraph. Whichgraph?
solution Averagevelocity is theslopeof asecant linethroughtwopoints onthegraphof positionas afunctionof
time.
2. Caninstantaneousvelocitybedeﬁnedasaratio? If not, howisinstantaneousvelocitycomputed?
solution Instantaneous velocity cannot bedeﬁned as aratio. It is deﬁned as thelimit of averagevelocity as time
elapsedshrinkstozero.
3. What isthegraphical interpretationof instantaneousvelocityat amoment t = t
0
?
solution Instantaneousvelocityat timet = t
0
istheslopeof thelinetangent tothegraphof positionasafunctionof
timeat t = t
0
.
4. What isthegraphical interpretationof thefollowingstatement?Theaveragerateof changeapproachestheinstanta-
neousrateof changeastheinterval [x
0
, x
1
] shrinkstox
0
.
solution Theslopeof thesecant lineover theinterval [x
0
, x
1
] approachestheslopeof thetangent lineat x = x
0
.
5. Therateof changeof atmospheric temperaturewithrespect toaltitudeisequal totheslopeof thetangent linetoa
graph. Whichgraph?What arepossibleunitsfor thisrate?
solution Therateof changeof atmospherictemperaturewithrespect toaltitudeistheslopeof thelinetangent tothe
graphof atmospherictemperatureasafunctionof altitude. Possibleunitsfor thisrateof changeare

F/ft or

C/m.
Exercises
1. A ball droppedfromastateof rest at timet = 0travelsadistances(t ) = 4.9t
2
mint seconds.
(a) Howfar doestheball travel duringthetimeinterval [2, 2.5]?
(b) Computetheaveragevelocityover [2, 2.5].
(c) Computetheaveragevelocityforthetimeintervalsinthetableandestimatetheball’sinstantaneousvelocityatt = 2.
Interval [2, 2.01] [2, 2.005] [2, 2.001] [2, 2.00001]
Average
velocity
solution
(a) Duringthetimeinterval [2, 2.5], theball travelss = s(2.5) −s(2) = 4.9(2.5)
2
−4.9(2)
2
= 11.025m.
(b) Theaveragevelocityover [2, 2.5] is
s
t
=
s(2.5) −s(2)
2.5−2
=
11.025
0.5
= 22.05m/s.
(c)
timeinterval [2, 2.01] [2, 2.005] [2, 2.001] [2, 2.00001]
averagevelocity 19.649 19.6245 19.6049 19.600049
Theinstantaneousvelocityat t = 2is19.6m/s.
A wrenchreleasedfromastateof rest at timet = 0travelsadistances(t ) = 4.9t
2
mint seconds. Estimatethe
instantaneousvelocityat t = 3.
3. Let v = 20

T asinExample2. Estimatetheinstantaneousrateof changeof v withrespect toT whenT = 300K.
solution
T interval [300, 300.01] [300, 300.005]
averagerateof change 0.577345 0.577348
T interval [300, 300.001] [300, 300.00001]
averagerateof change 0.57735 0.57735
Theinstantaneousrateof changeisapproximately0.57735m/(s· K).
31
June 7, 2011 LTSV SSM Second Pass
32 C HA P T E R 2 LIMITS
Computey/x for theinterval [2, 5], wherey = 4x − 9. What istheinstantaneousrateof changeof y with
respect tox at x = 2?
InExercises5and6, astoneistossedverticallyintotheair fromgroundlevel withaninitial velocityof 15m/s. Itsheight
at timet ish(t ) = 15t −4.9t
2
m.
5. Computethestone’saveragevelocityover thetimeinterval [0.5, 2.5] andindicatethecorrespondingsecant lineon
asketchof thegraphof h(t ).
solution Theaveragevelocityisequal to
h(2.5) −h(0.5)
2
= 0.3.
Thesecant lineisplottedwithh(t ) below.
2
0.5 1 1.5 2 2.5 3
4
6
8
10
t
h
Compute the stone’s average velocity over the time intervals [1, 1.01], [1, 1.001], [1, 1.0001] and [0.99, 1],
[0.999, 1], [0.9999, 1], andthenestimatetheinstantaneousvelocityat t = 1.
7. Withaninitial deposit of \$100, thebalanceinabankaccount after t yearsisf (t ) = 100(1.08)
t
dollars.
(a) What aretheunitsof therateof changeof f (t )?
(b) Findtheaveragerateof changeover [0, 0.5] and[0, 1].
(c) Estimatetheinstantaneousrateof changeat t = 0.5by computingtheaveragerateof changeover intervalstothe
left andright of t = 0.5.
solution
(a) Theunitsof therateof changeof f (t ) aredollars/year or \$/yr.
(b) Theaveragerateof changeof f (t ) = 100(1.08)
t
over thetimeinterval [t
1
, t
2
] isgivenby
f
t
=
f (t
2
) −f (t
1
)
t
2
−t
1
.
timeinterval [0, .5] [0, 1]
averagerateof change 7.8461 8
(c)
timeinterval [0.5, 0.51] [0.5, 0.501] [0.5, 0.5001]
averagerateof change 8.0011 7.9983 7.9981
timeinterval [0.49, 0.5] [0.499, 0.5] [0.4999, 0.5]
averagerateof change 7.9949 7.9977 7.998
Therateof changeat t = 0.5isapproximately\$8/yr.
Thepositionof aparticleat timet is s(t ) = t
3
+ t . Computetheaveragevelocity over thetimeinterval [1, 4]
andestimatetheinstantaneousvelocityat t = 1.
9. Figure8shows theestimatednumber N of Internet users inChile, basedondatafromtheUnitedNations
StatisticsDivision.
(a) Estimatetherateof changeof N at t = 2003.5.
(b) Doestherateof changeincreaseor decreaseast increases? Explaingraphically.
(c) Let R betheaveragerateof changeover [2001, 2005]. ComputeR.
(d) Istherateof changeat t = 2002greater thanor lessthantheaveragerateR? Explaingraphically.
2001 2002 2003 2004 2005
3.5
4.0
4.5
N (Internet users in Chilein millions)
t (years)
FIGURE 8
June 7, 2011 LTSV SSM Second Pass
S E C T I ON 2.1 Limits, Rates of Change, and Tangent Lines 33
solution
(a) Thetangent lineshowninFigure8appearstopassthroughthepoints(2002, 3.75) and(2005, 4.6). Thus, therateof
changeof N at t = 2003.5isapproximately
4.6−3.75
2005−2002
= 0.283
millionInternet usersper year.
(b) Ast increases, wemovefromlefttorightalongthegraphinFigure8. Moreover, aswemovefromlefttorightalong
thegraph, theslopeof thetangent linedecreases. Thus, therateof changedecreasesast increases.
(c) Thegraphof N(t ) appear topass throughthepoints (2001, 3.1) and(2005, 4.5). Thus, theaveragerateof change
over [2001, 2005] isapproximately
R =
4.5−3.1
2005−2001
= 0.35
millionInternet usersper year.
(d) For theﬁgurebelow, weseethat theslopeof thetangent lineat t = 2002islarger thantheslopeof thesecant line
through theendpoints of thegraph of N(t ). Thus, therateof changeat t = 2002 is greater than theaveragerateof
changeR.
3.0
2001 2002 2003 2004 2005
x
y
3.5
4.0
4.5
TheatmospherictemperatureT (in

C) ataltitudeh metersaboveacertainpointonearthisT = 15−0.0065h
for h ≤ 12,000m. What aretheaverageandinstantaneousratesof changeof T withrespect toh?Whyaretheythe
same? Sketchthegraphof T for h ≤ 12,000.
InExercises11–18, estimatetheinstantaneousrateof changeat thepoint indicated.
11. P(x) = 3x
2
−5; x = 2
solution
x interval [2, 2.01] [2, 2.001] [2, 2.0001] [1.99, 2] [1.999, 2] [1.9999, 2]
averagerateof change 12.03 12.003 12.0003 11.97 11.997 11.9997
Therateof changeat x = 2isapproximately12.
f (t ) = 12t −7; t = −4 13. y(x) =
1
x +2
; x = 2
solution
x interval [2, 2.01] [2, 2.001] [2, 2.0001] [1.99, 2] [1.999, 2] [1.9999, 2]
averagerateof change −0.0623 −0.0625 −0.0625 −0.0627 −0.0625 −0.0625
Therateof changeat x = 2isapproximately−0.06.
y(t ) =

3t +1; t = 1
15. f (x) = 3
x
; x = 0
solution
x interval [−0.01, 0] [−0.001, 0] [−0.0001, 0] [0, 0.01] [0, 0.001] [0, 0.0001]
averagerateof change 1.0926 1.098 1.0986 1.1047 1.0992 1.0987
Therateof changeisbetwenn1.0986and1.0987.
f (x) = 3
x
; x = 3
17. f (x) = sinx; x =
π
6
solution
x interval
_
π
6
−0.01,
π
6
_ _
π
6
−0.001,
π
6
_ _
π
6
−0.0001,
π
6
_ _
π
6
,
π
6
+0.01
_ _
π
6
,
π
6
+0.001
_ _
π
6
,
π
6
+0.001
_
averagerateof change 0.8685 0.8663 0.8660 0.8635 0.8658 0.8660
Therateof changeat x =
π
6
isapproximately0.866.
June 7, 2011 LTSV SSM Second Pass
34 C HA P T E R 2 LIMITS
f (x) = tanx; x =
π
4
19. Theheight(incentimeters)attimet (inseconds)of asmall massoscillatingattheendof aspringish(t ) = 8cos(12πt ).
(a) Calculatethemass’saveragevelocityover thetimeintervals[0, 0.1] and[3, 3.5].
(b) Estimateitsinstantaneousvelocityat t = 3.
solution
(a) Theaveragevelocityover thetimeinterval [t
1
, t
2
] isgivenby
h
t
=
h(t
2
) −h(t
1
)
t
2
−t
1
.
timeinterval [0, 0.1] [3, 3.5]
averagevelocity −144.721cm/s 0cm/s
(b)
timeinterval [3, 3.0001] [3, 3.00001] [3, 3.000001] [2.9999, 3] [2.99999, 3] [2.999999, 3]
averagevelocity −0.5685 −0.05685 −0.005685 0.5685 0.05685 0.005685
Theinstantaneousvelocityat t = 3secondsisapproximately0cm/s.
Thenumber P(t ) of E. coli cellsat timet (hours) inapetri dishisplottedinFigure9.
(a) Calculatetheaveragerateof changeof P(t ) over thetimeinterval [1, 3] anddrawthecorrespondingsecantline.
(b) Estimatetheslopem of thelineinFigure9. What doesm represent?
21. AssumethattheperiodT (inseconds) of apendulum(thetimerequiredfor acompleteback-and-forthcycle)
isT =
3
2

L, whereL isthependulum’slength(inmeters).
(a) What aretheunitsfor therateof changeof T withrespect toL? Explainwhat thisratemeasures.
(b) Whichquantitiesarerepresentedbytheslopesof linesA andB inFigure10?
(c) Estimatetheinstantaneousrateof changeof T withrespect toL whenL = 3m.
Period (s)
Length (m)
1 3
A
B
2
FIGURE 10 TheperiodT isthetimerequiredfor apendulumtoswingbackandforth.
solution
(a) Theunitsfor therateof changeof T withrespect toL aresecondsper meter. Thisratemeasuresthesensitivity of
theperiodof thependulumtoachangeinthelengthof thependulum.
(b) Theslopeof thelineB representstheaveragerateof changeinT fromL = 1mtoL = 3m. Theslopeof theline
A representstheinstantaneousrateof changeof T at L = 3m.
(c)
timeinterval [3, 3.01] [3, 3.001] [3, 3.0001] [2.99, 3] [2.999, 3] [2.9999, 3]
averagevelocity 0.4327 0.4330 0.4330 0.4334 0.4330 0.4330
Theinstantaneousrateof changeat L = 1misapproximately0.4330s/m.
ThegraphsinFigure11represent thepositionsof movingparticlesasfunctionsof time.
(a) Dotheinstantaneousvelocitiesat timest
1
, t
2
, t
3
in(A) formanincreasingor adecreasingsequence?
(b) Istheparticlespeedingupor slowingdownin(A)?
(c) Istheparticlespeedingupor slowingdownin(B)?
23. Anadvertisingcampaignboostedsalesof CrunchyCrustfrozenpizzatoapeaklevel of S
0
dollarsper month.
A marketingstudyshowedthat after t months, monthlysalesdeclinedto
S(t ) = S
0
g(t ), whereg(t ) =
1

1+t
.
Do sales declinemoreslowly or morerapidly as timeincreases? Answer by referringto asketchof thegraphof g(t )
together withseveral tangent lines.
solution Wenoticefromtheﬁgurebelowthat, astimeincreases, theslopesof thetangent linestothegraphof g(t )
becomelessnegative. Thus, salesdeclinemoreslowlyastimeincreases.
2
0.2
0.4
0.6
0.8
1.0
y
x
4 6 8 10 12
June 7, 2011 LTSV SSM Second Pass
S E C T I ON 2.1 Limits, Rates of Change, and Tangent Lines 35
Thefractionof acity’spopulationinfectedbyaﬂuvirusisplottedasafunctionof time(inweeks) inFigure12.
(a) Whichquantitiesarerepresentedbytheslopesof linesA andB? Estimatetheseslopes.
(b) Istheﬂuspreadingmorerapidlyat t = 1, 2, or 3?
(c) Istheﬂuspreadingmorerapidlyat t = 4, 5, or 6?
25. ThegraphsinFigure13represent thepositionss of movingparticlesasfunctionsof timet . Matcheachgraphwith
adescription:
(a) Speedingup
(b) Speedingupandthenslowingdown
(c) Slowingdown
(d) Slowingdownandthenspeedingup
(B) (A) (D) (C)
t
s
t
s
t
s
t
s
FIGURE 13
solution When aparticleis speeding up over atimeinterval, its graph is bent upward over that interval. When a
particleisslowingdown, itsgraphisbent downwardover that interval. Accordingly,
• Ingraph(A), theparticleis(c) slowingdown.
• Ingraph(B), theparticleis(b) speedingupandthenslowingdown.
• Ingraph(C), theparticleis(d) slowingdownandthenspeedingup.
• Ingraph(D), theparticleis(a) speedingup.
Anepidemiologist ﬁndsthat thepercentageN(t ) of susceptiblechildrenwhowereinfectedonday t duringthe
ﬁrst threeweeksof ameaslesoutbreakisgiven, toareasonableapproximation, bytheformula(Figure14)
N(t ) =
100t
2
t
3
+5t
2
−100t +380
(a) Drawthesecant linewhoseslopeistheaveragerateof changeininfectedchildrenover theintervals[4, 6] and
[12, 14]. Thencomputetheseaveragerates(inunitsof percent per day).
(b) Istherateof declinegreater at t = 8or t = 16?
(c) Estimatetherateof changeof N(t ) onday12.
27. ThefungusFusariumexosporiuminfectsaﬁeldof ﬂaxplantsthroughtherootsandcausestheplantstowilt.Eventually,
theentireﬁeldisinfected. Thepercentagef (t ) of infectedplantsasafunctionof timet (indays) sinceplantingisshown
inFigure15.
(a) What aretheunitsof therateof changeof f (t ) withrespect tot ?What doesthisratemeasure?
(b) Usethegraphto rank (fromsmallest to largest) theaverageinfectionrates over theintervals [0, 12], [20, 32], and
[40, 52].
(c) Usethefollowingtabletocomputetheaverageratesof infectionover theintervals[30, 40], [40, 50], [30, 50].
Days 0 10 20 30 40 50 60
Percent infected 0 18 56 82 91 96 98
(d) Drawthetangent lineat t = 40andestimateitsslope.
Percent infected
Days after planting
10 20 30 40 50 60
100
80
60
40
20
FIGURE 15
solution
(a) Theunitsof therateof changeof f (t ) withrespecttot arepercent/dayor %/d. Thisratemeasureshowquicklythe
populationof ﬂaxplantsisbecominginfected.
(b) Fromsmallest tolargest, theaverageratesof infectionarethoseover theintervals[40, 52], [0, 12], [20, 32]. Thisis
becausetheslopesof thesecant linesover theseintervalsarearrangedfromsmallest tolargest.
(c) Theaverageratesof infectionover theintervals[30, 40], [40, 50], [30, 50] are0.9, 0.5, 0.7%/d, respectively.
(d) Thetangent linesketchedinthegraphbelowappearstopassthroughthepoints(20, 80) and(40, 91). Theestimate
of theinstantaneousrateof infectionat t = 40daysistherefore
91−80
40−20
=
11
20
= 0.55%/d.
June 7, 2011 LTSV SSM Second Pass
36 C HA P T E R 2 LIMITS
10 20 30 40 50 60
100
80
60
40
20
Letv = 20

T asinExample2. Istherateof changeof v withrespecttoT greater atlowtemperaturesor high
temperatures? Explainintermsof thegraph.
29. If anobject inlinear motion(but withchangingvelocity) coverss metersint seconds, thenitsaverage
velocity is v
0
= s/t m/s. Showthat it wouldcover thesamedistanceif it traveledat constant velocity v
0
over the
sametimeinterval. Thisjustiﬁesour callings/t theaveragevelocity.
solution At constant velocity, thedistancetraveledisequal tovelocity timestime, soanobject movingat constant
velocityv
0
for t secondstravelsv
0
δt meters. Sincev
0
= s/t , weﬁnd
distancetraveled= v
0
δt =
_
s
t
_
t = s
Sotheobject coversthesamedistances bytravelingat constant velocityv
0
.
Sketchthegraphof f (x) = x(1− x) over [0, 1]. Refer tothegraphand, without makingany computations,
ﬁnd:
(a) Theaveragerateof changeover [0, 1]
(b) The(instantaneous) rateof changeat x =
1
2
(c) Thevaluesof x at whichtherateof changeispositive
31. WhichgraphinFigure16has thefollowingproperty: For all x, theaveragerateof changeover [0, x] is
greater thantheinstantaneousrateof changeat x. Explain.
(B)
x
y
(A)
x
y
FIGURE 16
solution
(a) Theaveragerateof changeover [0, x] isgreater thantheinstantaneousrateof changeat x: (B).
(b) Theaveragerateof changeover [0, x] islessthantheinstantaneousrateof changeat x: (A)
Thegraphin(B) bends downward, so theslopeof thesecant linethrough(0, 0) and(x, f (x)) is larger thantheslope
of thetangent lineat (x, f (x)). Ontheother hand, thegraphin(A) bends upward, so theslopeof thetangent lineat
(x, f (x)) islarger thantheslopeof thesecant linethrough(0, 0) and(x, f (x)).
Further Insights and Challenges
Theheight of aprojectileﬁredintheair verticallywithinitial velocity25m/sis
h(t ) = 25t −4.9t
2
m.
(a) Computeh(1). Showthat h(t ) −h(1) canbefactoredwith(t −1) asafactor.
(b) Usingpart (a), showthat theaveragevelocityover theinterval [1, t ] is20.1−4.9t .
(c) Usethis formulato ﬁndtheaveragevelocity over several intervals [1, t ] witht closeto 1. Thenestimatethe
instantaneousvelocityat timet = 1.
33. Let Q(t ) = t
2
. Asinthepreviousexercise, ﬁndaformulafor theaveragerateof changeof Qover theinterval [1, t ]
anduseit to estimatetheinstantaneous rateof changeat t = 1. Repeat for theinterval [2, t ] andestimatetherateof
changeat t = 2.
solution Theaveragerateof changeis
Q(t ) −Q(1)
t −1
=
t
2
−1
t −1
.
Applyingthedifferenceof squaresformulagivesthattheaveragerateof changeis((t +1)(t −1))/(t −1) = (t +1) for
t = 1. Ast getscloser to1, thisgetscloser to1+1= 2. Theinstantaneousrateof changeis2.
For t
0
= 2, theaveragerateof changeis
Q(t ) −Q(2)
t −2
=
t
2
−4
t −2
,
whichsimpliﬁestot + 2for t = 2. Ast approaches2, theaveragerateof changeapproaches4. Theinstantaneousrate
of changeistherefore4.
Showthat theaveragerateof changeof f (x) = x
3
over [1, x] isequal to
x
2
+x +1.
Usethistoestimatetheinstantaneousrateof changeof f (x) at x = 1.
June 7, 2011 LTSV SSM Second Pass
S E C T I ON 2.2 Limits: A Numerical and Graphical Approach 37
35. Findaformulafor theaveragerateof changeof f (x) = x
3
over [2, x] anduseit toestimatetheinstantaneousrate
of changeat x = 2.
solution Theaveragerateof changeis
f (x) −f (2)
x −2
=
x
3
−8
x −2
.
Applyingthedifferenceof cubesformulatothenumerator, weﬁndthat theaveragerateof changeis
(x
2
+2x +4)(x −2)
x −2
= x
2
+2x +4
for x = 2. Thecloser x getsto2, thecloser theaveragerateof changegetsto2
2
+2(2) +4= 12.
LetT =
3
2

L asinExercise21. Thenumbersinthesecondcolumnof Table4areincreasing, andthoseinthe
lastcolumnaredecreasing. Explainwhyintermsof thegraphof T asafunctionof L. Also, explaingraphicallywhy
theinstantaneousrateof changeat L = 3liesbetween0.4329and0.4331.
2.2 Limits: A Numerical and Graphical Approach
Preliminary Questions
1. What isthelimit of f (x) = 1asx → π?
solution lim
x→π
1= 1.
2. What isthelimit of g(t ) = t ast → π?
solution lim
t →π
t = π.
3. Is lim
x→10
20equal to10or 20?
solution lim
x→10
20= 20.
4. Canf (x) approachalimit asx → c if f (c) isundeﬁned? If so, giveanexample.
solution Yes. Thelimitof afunctionf asx → c doesnotdependonwhathappensatx = c, onlyonthebehavior of
f asx → c. Asanexample, consider thefunction
f (x) =
x
2
−1
x −1
.
Thefunctionisclearlynot deﬁnedat x = 1but
lim
x→1
f (x) = lim
x→1
x
2
−1
x −1
= lim
x→1
(x +1) = 2.
5. What doesthefollowingtablesuggest about lim
x→1−
f (x) and lim
x→1+
f (x)?
x 0.9 0.99 0.999 1.1 1.01 1.001
f (x) 7 25 4317 3.0126 3.0047 3.00011
solution Thevaluesinthetablesuggest that lim
x→1−
f (x) = ∞andlim
x→1+
f (x) = 3.
6. Canyoutell whether lim
x→5
f (x) existsfromaplot of f (x) for x > 5? Explain.
solution No. Byexaminingvaluesof f (x) for x closetobut greater than5, wecandeterminewhether theone-sided
limit lim
x→5+
f (x) exists. Todeterminewhether lim
x→5
f (x) exists, wemust examinevalueof f (x) onbothsidesof
x = 5.
7. If youknowinadvancethat lim
x→5
f (x) exists, canyoudetermineitsvaluefromaplot of f (x) for all x > 5?
solution Yes. If lim
x→5
f (x) exists, thenbothone-sidedlimitsmust exist andbeequal.
June 7, 2011 LTSV SSM Second Pass
38 C HA P T E R 2 LIMITS
Exercises
InExercises1–4, ﬁll inthetablesandguessthevalueof thelimit.
1. lim
x→1
f (x), wheref (x) =
x
3
−1
x
2
−1
.
x f (x) x f (x)
1.002 0.998
1.001 0.999
1.0005 0.9995
1.00001 0.99999
solution
x 0.998 0.999 0.9995 0.99999 1.00001 1.0005 1.001 1.002
f (x) 1.498501 1.499250 1.499625 1.499993 1.500008 1.500375 1.500750 1.501500
Thelimit asx → 1is
3
2
.
lim
t →0
h(t ), whereh(t ) =
cost −1
t
2
. Notethat h(t ) iseven; that is, h(t ) = h(−t ).
t ±0.002 ±0.0001 ±0.00005 ±0.00001
h(t )
3. lim
y→2
f (y), wheref (y) =
y
2
−y −2
y
2
+y −6
.
y f (y) y f (y)
2.002 1.998
2.001 1.999
2.0001 1.9999
solution
y 1.998 1.999 1.9999 2.0001 2.001 2.02
f (y) 0.59984 0.59992 0.599992 0.600008 0.60008 0.601594
Thelimit asy → 2is
3
5
.
lim
θ→0
f (θ), wheref (θ) =
sinθ −θ
θ
3
.
θ ±0.002 ±0.0001 ±0.00005 ±0.00001
f (θ)
5. Determine lim
x→0.5
f (x) for f (x) asinFigure9.
0.5
1.5
x
y
1
f(x)
FIGURE 9
solution Thegraphsuggeststhat f (x) → 1.5asx → 0.5.
Determine lim
x→0.5
g(x) for g(x) asinFigure10.
InExercises7and8, evaluatethelimit.
7. lim
x→21
x
solution Asx → 21, f (x) = x → 21. Youcanseethis, for example, onthegraphof f (x) = x.
lim
x→4.2

3
InExercises 9–16, verifyeachlimit usingthelimit deﬁnition. For example, inExercise9, showthat |3x − 12| canbe
madeassmall asdesiredbytakingx closeto4.
9. lim
x→4
3x = 12
solution |3x −12| = 3|x −4|. |3x −12| canbemadearbitrarilysmall bymakingx closeenoughto4, thusmaking
|x −4| small.
June 7, 2011 LTSV SSM Second Pass
S E C T I ON 2.2 Limits: A Numerical and Graphical Approach 39
lim
x→5
3= 3
11. lim
x→3
(5x +2) = 17
solution |(5x + 2) − 17| = |5x − 15| = 5|x − 3|. Therefore, if you make|x − 3| small enough, you can make
|(5x +2) −17| assmall asdesired.
lim
x→2
(7x −4) = 10
13. lim
x→0
x
2
= 0
solution Asx → 0, wehave|x
2
−0| = |x +0||x −0|. Tosimplifythings, supposethat |x| < 1, sothat |x +0||x −
0| = |x||x| < |x|. Bymaking|x| sufﬁcientlysmall, sothat |x +0||x −0| = x
2
isevensmaller, youcanmake|x
2
−0|
assmall asdesired.
lim
x→0
(3x
2
−9) = −9
15. lim
x→0
(4x
2
+2x +5) = 5
solution Asx → 0, wehave|4x
2
+2x +5−5| = |4x
2
+2x| = |x||4x +2|. If |x| < 1, |4x +2| canbenobigger
than6, so|x||4x +2| < 6|x|. Therefore, bymaking|x −0| = |x| sufﬁcientlysmall, youcanmake|4x
2
+2x +5−5| =
|x||4x +2| assmall asdesired.
lim
x→0
(x
3
+12) = 12
In Exercises 17–36, estimate the limit numerically or state that the limit does not exist. If inﬁnite, state whether the
one-sidedlimitsare∞or −∞.
17. lim
x→1

x −1
x −1
solution
x 0.9995 0.99999 1.00001 1.0005
f (x) 0.500063 0.500001 0.49999 0.499938
Thelimit asx → 1is
1
2
.
lim
x→−4
2x
2
−32
x +4
19. lim
x→2
x
2
+x −6
x
2
−x −2
solution
x 1.999 1.99999 2.00001 2.001
f (x) 1.666889 1.666669 1.666664 1.666445
Thelimit asx → 2is
5
3
.
lim
x→3
x
3
−2x
2
−9
x
2
−2x −3
21. lim
x→0
sin2x
x
solution
x −0.01 −0.005 0.005 0.01
f (x) 1.999867 1.999967 1.999967 1.999867
Thelimit asx → 0is2.
lim
x→0
sin5x
x
23. lim
θ→0
cosθ −1
θ
solution
x −0.05 −0.001 0.001 0.05
f (x) 0.0249948 0.0005 −0.0005 −0.0249948
Thelimit asx → 0is0.
lim
x→0
sinx
x
2
25. lim
x→4
1
(x −4)
3
solution
x 3.99 3.999 3.9999 4.0001 4.001 4.01
f (x) −10
6
−10
9
−10
12
10
12
10
9
10
6
Thelimit doesnot exist. Asx → 4−, f (x) → −∞; similarly, asx → 4+, f (x) → ∞.
June 7, 2011 LTSV SSM Second Pass
40 C HA P T E R 2 LIMITS
lim
x→1−
3−x
x −1
27. lim
x→3+
x −4
x
2
−9
solution
x 3.01 3.001 3.0001 3.00001
f (x) −16.473 −166.473 −1666.473 −16666.473
Asx → 3+, f (x) → −∞.
lim
h→0
3
h
−1
h
29. lim
h→0
sinhcos
1
h
solution
h −0.01 −0.001 −0.0001 0.0001 0.001 0.01
f (h) −0.008623 −0.000562 0.000095 −0.000095 0.000562 0.008623
Thelimit asx → 0is0.
lim
h→0
cos
1
h
31. lim
x→0
|x|
x
solution
x −0.05 −0.001 −0.00001 0.00001 0.001 0.05
f (x) 1.161586 1.006932 1.000115 0.999885 0.993116 0.860892
Thelimit asx → 0is1.
lim
x→0
2
x
−3
x
x
33. lim
θ→
π
4
tanθ −2sinθ cosθ
θ −
π
4
solution
θ
π
4
−0.01
π
4
−0.001
π
4
−0.0001
π
4
+0.0001
π
4
+0.001
π
4
+0.01
f (θ) 1.96026 1.99600 1.99960 2.00040 2.00400 2.04027
Thelimit asx →
π
4
isapproximately2.
lim
r→0
(1+r)
1/r 35. lim
θ→0
1−cosθ
θ
2
solution
θ −0.01 −0.001 −0.0001 0.0001 0.001 0.01
f (θ) 0.499996 0.500000 0.500000 0.500000 0.500000 0.499996
Thelimit asθ → 0appearstobe0.5.
lim
θ→0
1−cosθ
θ
3
37. Thegreatestinteger functionisdeﬁnedby[x] = n, wheren istheuniqueinteger suchthat n ≤ x < n +1. Sketch
thegraphof y = [x]. Calculate, for c aninteger:
(a) lim
x→c−
[x] (b) lim
x→c+
[x]
solution Hereisagraphof thegreatest integer function:
2
1
1 2 3 −1
x
y
(a) Fromthegraph, weseethat, for c aninteger,
lim
x→c−
[x] = c −1.
June 7, 2011 LTSV SSM Second Pass
S E C T I ON 2.2 Limits: A Numerical and Graphical Approach 41
(b) Fromthegraph, weseethat, for c aninteger,
lim
x→c+
[x] = c.
Determinetheone-sidedlimitsat c = 1, 2, and4of thefunctiong(x) showninFigure11, andstatewhether the
limit existsat thesepoints.
InExercises39–46, determinetheone-sidedlimitsnumericallyor graphically. Ifinﬁnite, statewhether theone-sidedlimits
are∞ or −∞, anddescribethecorrespondingvertical asymptote. InExercise46, [x] is thegreatest integer function
deﬁnedinExercise37.
39. lim
x→0±
sinx
|x|
solution
x −0.2 −0.02 0.02 0.2
f (x) −0.993347 −0.999933 0.999933 0.993347
Theleft-handlimit is lim
x→0−
f (x) = −1, whereastheright-handlimit is lim
x→0+
f (x) = 1.
lim
x→0±
|x|
1/x 41. lim
x→0±
x −sin|x|
x
3
solution
x −0.1 −0.01 0.01 0.1
f (x) 199.853 19999.8 0.166666 0.166583
Theleft-handlimitis lim
x→0−
f (x) = ∞, whereastheright-handlimitis lim
x→0+
f (x) =
1
6
. Thus, thelinex = 0isavertical
asymptotefromtheleft for thegraphof y =
x−sin|x|
x
3
.
lim
x→4±
x +1
x −4
43. lim
x→−2±
4x
2
+7
x
3
+8
solution Thegraphof y =
4x
2
+7
x
3
+8
for x near −2isshownbelow. Fromthisgraph, weseethat
lim
x→−2−
4x
2
+7
x
3
+8
= −∞ while lim
x→−2+
4x
2
+7
x
3
+8
= ∞.
Thus, thelinex = −2isavertical asymptotefor thegraphof y =
4x
2
+7
x
3
+8
.
−3.0 −2.5 −2.0 −1.5 −1.0
x
lim
x→−3±
x
2
x
2
−9
45. lim
x→1±
x
5
+x −2
x
2
+x −2
solution Thegraphof y =
x
5
+x−2
x
2
+x−2
for x near 1isshownbelow. Fromthisgraph, weseethat
lim
x→1±
x
5
+x −2
x
2
+x −2
= 2.
2
0.5 1.0 1.5
x
y
4
6
June 7, 2011 LTSV SSM Second Pass
42 C HA P T E R 2 LIMITS
lim
x→2±
cos
_
π
2
(x −[x])
_ 47. Determinetheone-sidedlimits at c = 2, 4of thefunctionf (x) inFigure12. What arethevertical asymptotes of
f (x)?
−5
4 2
15
5
10
x
y
FIGURE 12
solution
• For c = 2, wehave lim
x→2−
f (x) = ∞and lim
x→2+
f (x) = ∞.
• For c = 4, wehave lim
x→4−
f (x) = −∞and lim
x→4+
f (x) = 10.
Thevertical asymptotesarethevertical linesx = 2andx = 4.
Determinetheinﬁniteone- andtwo-sidedlimitsinFigure13.
InExercises49–52, sketchthegraphof afunctionwiththegivenlimits.
49. lim
x→1
f (x) = 2, lim
x→3−
f (x) = 0, lim
x→3+
f (x) = 4
solution
2
4
6
1 2 3 4
y
x
lim
x→1
f (x) = ∞, lim
x→3−
f (x) = 0, lim
x→3+
f (x) = −∞
51. lim
x→2+
f (x) = f (2) = 3, lim
x→2−
f (x) = −1, lim
x→4
f (x) = 2= f (4)
solution
1
−1
2
3
1 2 3 4 5
y
x
lim
x→1+
f (x) = ∞, lim
x→1−
f (x) = 3, lim
x→4
f (x) = −∞
53. Determinetheone-sidedlimitsof thefunctionf (x) inFigure14, at thepointsc = 1, 3, 5, 6.
−1
−2
−3
−4
1
2
3
4
5
y
x
1 2 3 4 5 6 7 8
FIGURE 14 Graphof f (x)
solution
• lim
x→1−
f (x) = lim
x→1+
f (x) = 3
• lim
x→3−
f (x) = −∞
June 7, 2011 LTSV SSM Second Pass
S E C T I ON 2.2 Limits: A Numerical and Graphical Approach 43
• lim
x→3+
f (x) = 4
• lim
x→5−
f (x) = 2
• lim
x→5+
f (x) = −3
• lim
x→6−
f (x) = lim
x→6+
f (x) = ∞
Doeseither of thetwooscillatingfunctionsinFigure15appear toapproachalimit asx → 0?
InExercises55–60, plot thefunctionandusethegraphtoestimatethevalueof thelimit.
55. lim
θ→0
sin5θ
sin2θ
solution
2.42
2.44
2.46
2.48
2.50
y
Fromthegraphof y =
sin5θ
sin2θ
shownabove, weseethat thelimit asθ → 0is
5
2
.
lim
x→0
12
x
−1
4
x
−1
57. lim
x→0
2
x
−cosx
x
solution
0.6935
0.6940
0.6930
0.6925
0.6920
y
y =
2
x
− cosx
x
Thelimit asx → 0isapproximately0.693. (Theexact answer isln2.)
lim
θ→0
sin
2

cosθ −1
59. lim
θ→0
cos7θ −cos5θ
θ
2
solution
−12.0
−11.8
−11.6
−11.4
y
Fromthegraphof y =
cos7θ −cos5θ
θ
2
shownabove, weseethat thelimit asθ → 0is−12.
lim
θ→0
sin
2
2θ −θ sin4θ
θ
4
61. Let n beapositiveinteger. For whichn arethetwoinﬁniteone-sidedlimits lim
x→0±
1/x
n
equal?
solution First, supposethat n iseven. Thenx
n
≥ 0for all x, and
1
x
n
> 0for all x = 0. Hence,
lim
x→0−
1
x
n
= lim
x→0+
1
x
n
= ∞.
Next, supposethat n isodd. Then
1
x
n
> 0for all x > 0but
1
x
n
< 0for all x < 0. Thus,
lim
x→0−
1
x
n
= −∞ but lim
x→0+
1
x
n
= ∞.
Finally, thetwoinﬁniteone-sidedlimitsareequal whenever n iseven.
Let L(n) = lim
x→1
_
n
1−x
n

1
1−x
_
for n apositiveinteger. InvestigateL(n) numericallyfor several valuesof
n, andthenguessthevalueof of L(n) ingeneral.
June 7, 2011 LTSV SSM Second Pass
44 C HA P T E R 2 LIMITS
63. Insomecases, numerical investigationscanbemisleading. Plot f (x) = cos
π
x
.
(a) Does lim
x→0
f (x) exist?
(b) Show, byevaluatingf (x) at x = ±
1
2
, ±
1
4
, ±
1
6
, . . . , that youmight beabletotrick your friendsintobelievingthat
thelimit existsandisequal toL = 1.
(c) Whichsequenceof evaluationsmight trickthemintobelievingthat thelimit isL = −1.
solution Hereisthegraphof f (x).
−0.05
−0.5
0.5
y
x
0.05
(a) Fromthegraphof f (x), whichshowsthat thevalueof f (x) oscillatesmoreandmorerapidlyasx → 0, it follows
that lim
x→0
f (x) doesnot exist.
(b) Noticethat
f
_
±
1
2
_
= cos
π
±1/2
= cos±2π = 1;
f
_
±
1
4
_
= cos
π
±1/4
= cos±4π = 1;
f
_
±
1
6
_
= cos
π
±1/6
= cos±6π = 1;
and, ingeneral, f (±
1
2n
) = 1for all integersn.
(c) At x = ±1, ±
1
3
, ±
1
5
, . . ., thevalueof f (x) isalways−1.
Further Insights and Challenges
Light wavesof frequencyλ passingthroughaslit of widtha produceaFraunhofer diffractionpatternof light
anddarkfringes(Figure16). Theintensityasafunctionof theangleθ is
I (θ) = I
m
_
sin(Rsinθ)
Rsinθ
_
2
whereR = πa/λ andI
m
isaconstant. Showthattheintensityfunctionisnotdeﬁnedatθ = 0. Thenchooseanytwo
valuesfor R andchecknumericallythat I (θ) approachesI
m
asθ → 0.
65. Investigate lim
θ→0
sinnθ
θ
numericallyfor several valuesof n. Thenguessthevalueingeneral.
solution
• For n = 3, wehave
θ −0.1 −0.01 −0.001 0.001 0.01 0.1
sinnθ
θ
2.955202 2.999550 2.999996 2.999996 2.999550 2.955202
Thelimit asθ → 0is3.
• For n = −5, wehave
θ −0.1 −0.01 −0.001 0.001 0.01 0.1
sinnθ
θ
−4.794255 −4.997917 −4.999979 −4.999979 −4.997917 −4.794255
Thelimit asθ → 0is−5.
• Wesurmisethat, ingeneral, lim
θ→0
sinnθ
θ
= n.
Shownumerically that lim
x→0
b
x
−1
x
for b = 3, 5appearstoequal ln3, ln5, wherelnx isthenatural logarithm.
Thenmakeaconjecture(guess) for thevalueingeneral andtest your conjecturefor twoadditional valuesof b.
67. Investigate lim
x→1
x
n
−1
x
m
−1
for (m, n) equal to (2, 1), (1, 2), (2, 3), and (3, 2). Then guess thevalueof thelimit in
general andcheckyour guessfor twoadditional pairs.
solution

x 0.99 0.9999 1.0001 1.01
x −1
x
2
−1
0.502513 0.500025 0.499975 0.497512
June 7, 2011 LTSV SSM Second Pass
S E C T I ON 2.2 Limits: A Numerical and Graphical Approach 45
Thelimit asx → 1is
1
2
.
x 0.99 0.9999 1.0001 1.01
x
2
−1
x −1
1.99 1.9999 2.0001 2.01
Thelimit asx → 1is2.
x 0.99 0.9999 1.0001 1.01
x
2
−1
x
3
−1
0.670011 0.666700 0.666633 0.663344
Thelimit asx → 1is
2
3
.
x 0.99 0.9999 1.0001 1.01
x
3
−1
x
2
−1
1.492513 1.499925 1.500075 1.507512
Thelimit asx → 1is
3
2
.
• For general m andn, wehave lim
x→1
x
n
−1
x
m
−1
=
n
m
.

x 0.99 0.9999 1.0001 1.01
x −1
x
3
−1
0.336689 0.333367 0.333300 0.330022
Thelimit asx → 1is
1
3
.
x 0.99 0.9999 1.0001 1.01
x
3
−1
x −1
2.9701 2.9997 3.0003 3.0301
Thelimit asx → 1is3.
x 0.99 0.9999 1.0001 1.01
x
3
−1
x
7
−1
0.437200 0.428657 0.428486 0.420058
Thelimit asx → 1is
3
7
≈ 0.428571.
Findbynumerical experimentationthepositiveintegersk suchthat lim
x→0
sin(sin
2
x)
x
k
exists.
69. Plot thegraphof f (x) =
2
x
−8
x −3
.
(a) ZoominonthegraphtoestimateL = lim
x→3
f (x).
(b) Explainwhy
f (2.99999) ≤ L ≤ f (3.00001)
UsethistodetermineL tothreedecimal places.
solution
(a)
5.555
5.565
5.545
5.535
5.525
y
x = 3
y =
2
x
− 8
x− 3
June 7, 2011 LTSV SSM Second Pass
46 C HA P T E R 2 LIMITS
(b) It is clear that thegraphof f rises as wemoveto theright. Mathematically, wemay express this observationas:
whenever u < v, f (u) < f (v). Because
2.99999< 3= lim
x→3
f (x) < 3.00001,
it followsthat
f (2.99999) < L = lim
x→3
f (x) < f (3.00001).
Withf (2.99999) ≈ 5.54516andf (3.00001) ≈ 5.545195, theaboveinequality becomes 5.54516 < L < 5.545195;
hence, tothreedecimal places, L = 5.545.
Thefunctionf (x) =
2
1/x
−2
−1/x
2
1/x
+2
−1/x
isdeﬁnedfor x = 0.
(a) Investigate lim
x→0+
f (x) and lim
x→0−
f (x) numerically.
(b) Plot thegraphof f anddescribeitsbehavior near x = 0.
2.3 Basic Limit Laws
Preliminary Questions
1. StatetheSumLawandQuotient Law.
solution Supposelim
x→c
f (x) andlim
x→c
g(x) bothexist. TheSumLawstatesthat
lim
x→c
(f (x) +g(x)) = lim
x→c
f (x) + lim
x→c
g(x).
Providedlim
x→c
g(x) = 0, theQuotient Lawstatesthat
lim
x→c
f (x)
g(x)
=
lim
x→c
f (x)
lim
x→c
g(x)
.
2. Whichof thefollowingisaverbal versionof theProduct Law(assumingthelimitsexist)?
(a) Theproduct of twofunctionshasalimit.
(b) Thelimit of theproduct istheproduct of thelimits.
(c) Theproduct of alimit isaproduct of functions.
(d) A limit producesaproduct of functions.
solution Theverbal versionof theProduct Lawis(b): Thelimit of theproduct istheproduct of thelimits.
3. Whichstatement iscorrect?TheQuotient Lawdoesnot holdif:
(a) Thelimit of thedenominator iszero.
(b) Thelimit of thenumerator iszero.
solution Statements(a)iscorrect. TheQuotient Lawdoesnot holdif thelimit of thedenominator iszero.
Exercises
InExercises1–24, evaluatethelimit usingtheBasicLimit Lawsandthelimits lim
x→c
x
p/q
= c
p/q
and lim
x→c
k = k.
1. lim
x→9
x
solution lim
x→9
x = 9.
lim
x→−3
14
3. lim
x→
1
2
x
4
solution lim
x→
1
2
x
4
=
_
1
2
_
4
=
1
16
.
lim
z→27
z
2/3
5. lim
t →2
t
−1
solution lim
t →2
t
−1
= 2
−1
=
1
2
.
lim
x→5
x
−2
7. lim
x→0.2
(3x +4)
solution UsingtheSumLawandtheConstant MultipleLaw:
lim
x→0.2
(3x +4) = lim
x→0.2
3x + lim
x→0.2
4
= 3 lim
x→0.2
x + lim
x→0.2
4= 3(0.2) +4= 4.6.
June 7, 2011 LTSV SSM Second Pass
S E C T I ON 2.3 Basic Limit Laws 47
lim
x→
1
3
(3x
3
+2x
2
)
9. lim
x→−1
(3x
4
−2x
3
+4x)
solution UsingtheSumLaw, theConstant MultipleLawandthePowersLaw:
lim
x→−1
(3x
4
−2x
3
+4x) = lim
x→−1
3x
4
− lim
x→−1
2x
3
+ lim
x→−1
4x
= 3 lim
x→−1
x
4
−2 lim
x→−1
x
3
+4 lim
x→−1
x
= 3(−1)
4
−2(−1)
3
+4(−1) = 3+2−4= 1.
lim
x→8
(3x
2/3
−16x
−1
)
11. lim
x→2
(x +1)(3x
2
−9)
solution UsingtheProduct Law, theSumLawandtheConstant MultipleLaw:
lim
x→2
(x +1)
_
3x
2
−9
_
=
_
lim
x→2
x + lim
x→2
1
__
lim
x→2
3x
2
− lim
x→2
9
_
= (2+1)
_
3 lim
x→2
x
2
−9
_
= 3(3(2)
2
−9) = 9.
lim
x→
1
2
(4x +1)(6x −1) 13. lim
t →4
3t −14
t +1
solution UsingtheQuotient Law, theSumLawandtheConstant MultipleLaw:
lim
t →4
3t −14
t +1
=
lim
t →4
(3t −14)
lim
t →4
(t +1)
=
3lim
t →4
t − lim
t →4
14
lim
t →4
t + lim
t →4
1
=
3· 4−14
4+1
= −
2
5
.
lim
z→9

z
z −2
15. lim
y→
1
4
(16y +1)(2y
1/2
+1)
solution UsingtheProduct Law, theSumLaw, theConstant MultipleLawandthePowersLaw:
lim
y→
1
4
(16y +1)(2y
1/2
+1) =
_
lim
y→
1
4
(16y +1)
__
lim
y→
1
4
(2y
1/2
+1)
_
=
_
16 lim
y→
1
4
y + lim
y→
1
4
1
__
2 lim
y→
1
4
y
1/2
+ lim
y→
1
4
1
_
=
_
16
_
1
4
_
+1
__
2
_
1
2
_
+1
_
= 10.
lim
x→2
x(x +1)(x +2) 17. lim
y→4
1

6y +1
solution UsingtheQuotient Law, thePowersLaw, theSumLawandtheConstant MultipleLaw:
lim
y→4
1

6y +1
=
1
lim
y→4

6y +1
=
1
_
6 lim
y→4
y +1
=
1

6(4) +1
=
1
5
.
lim
w→7

w +2+1

w −3−1
19. lim
x→−1
x
x
3
+4x
solution UsingtheQuotient Law, theSumLaw, thePowersLawandtheConstant MultipleLaw:
lim
x→−1
x
x
3
+4x
=
lim
x→−1
x
lim
x→−1
x
3
+4 lim
x→−1
x
=
−1
(−1)
3
+4(−1)
=
1
5
.
lim
t →−1
t
2
+1
(t
3
+2)(t
4
+1)
June 7, 2011 LTSV SSM Second Pass
48 C HA P T E R 2 LIMITS
21. lim
t →25
3

t −
1
5
t
(t −20)
2
solution UsingtheQuotient Law, theSumLaw, theConstant MultipleLawandthePowersLaw:
lim
t →25
3

t −
1
5
t
(t −20)
2
=
3
_
lim
t →25
t −
1
5
lim
t →25
t
_
lim
t →25
t −20
_
2
=
3(5) −
1
5
(25)
5
2
=
2
5
.
lim
y→
1
3
(18y
2
−4)
4
23. lim
t →
3
2
(4t
2
+8t −5)
3/2
solution UsingthePowersLaw, theSumLawandtheConstant MultipleLaw:
lim
t →
3
2
(4t
2
+8t −5)
3/2
=
_
4 lim
t →
3
2
t
2
+8 lim
t →
3
2
t −5
_
3/2
= (9+12−5)
3/2
= 64.
lim
t →7
(t +2)
1/2
(t +1)
2/3
25. UsetheQuotient Lawtoprovethat if lim
x→c
f (x) existsandisnonzero, then
lim
x→c
1
f (x)
=
1
lim
x→c
f (x)
solution Since lim
x→c
f (x) isnonzero, wecanapplytheQuotient Law:
lim
x→c
_
1
f (x)
_
=
_
lim
x→c
1
_
_
lim
x→c
f (x)
_ =
1
lim
x→c
f (x)
.
Assumingthat lim
x→6
f (x) = 4, compute:
(a) lim
x→6
f (x)
2
(b) lim
x→6
1
f (x)
(c) lim
x→6
x
_
f (x)
InExercises27–30, evaluatethelimit assumingthat lim
x→−4
f (x) = 3and lim
x→−4
g(x) = 1.
27. lim
x→−4
f (x)g(x)
solution lim
x→−4
f (x)g(x) = lim
x→−4
f (x) lim
x→−4
g(x) = 3· 1= 3.
lim
x→−4
(2f (x) +3g(x)) 29. lim
x→−4
g(x)
x
2
solution Since lim
x→−4
x
2
= 0, wemayapplytheQuotient Law, thenapplyingthePowersLaw:
lim
x→−4
g(x)
x
2
=
lim
x→−4
g(x)
lim
x→−4
x
2
=
1
_
lim
x→−4
x
_
2
=
1
16
.
lim
x→−4
f (x) +1
3g(x) −9
31. CantheQuotient Lawbeappliedtoevaluate lim
x→0
sinx
x
? Explain.
solution ThelimitQuotientLawcannot beappliedtoevaluate lim
x→0
sinx
x
since lim
x→0
x = 0. Thisviolatesacondition
of theQuotient Law. Accordingly, therulecannot beemployed.
Showthat theProduct Lawcannot beusedtoevaluatethelimit lim
x→π/2
_
x −
π
2
_
tanx.
33. Giveanexamplewhere lim
x→0
(f (x) +g(x)) existsbut neither lim
x→0
f (x) nor lim
x→0
g(x) exists.
solution Let f (x) = 1/x and g(x) = −1/x. Then lim
x→0
(f (x) +g(x)) = lim
x→0
0= 0. However, lim
x→0
f (x) =
lim
x→0
1/x and lim
x→0
g(x) = lim
x→0
−1/x donot exist.
June 7, 2011 LTSV SSM Second Pass
S E C T I ON 2.4 Limits and Continuity 49
Further Insights and Challenges
Show that if both lim
x→c
f (x) g(x) and lim
x→c
g(x) exist and lim
x→c
g(x) = 0, then lim
x→c
f (x) exists. Hint: Write
f (x) =
f (x) g(x)
g(x)
.
35. Supposethat lim
t →3
tg(t ) = 12. Showthat lim
t →3
g(t ) existsandequals4.
solution Wearegiventhat lim
t →3
tg(t ) = 12. Since lim
t →3
t = 3= 0, wemayapplytheQuotient Law:
lim
t →3
g(t ) = lim
t →3
tg(t )
t
=
lim
t →3
tg(t )
lim
t →3
t
=
12
3
= 4.
Provethat if lim
t →3
h(t )
t
= 5, then lim
t →3
h(t ) = 15.
37. Assumingthat lim
x→0
f (x)
x
= 1, whichof thefollowingstatementsisnecessarilytrue?Why?
(a) f (0) = 0 (b) lim
x→0
f (x) = 0
solution
(a) Given that lim
x→0
f (x)
x
= 1, it is not necessarily true that f (0) = 0. A counterexample is provided by f (x) =
_
x, x = 0
5, x = 0
.
(b) Giventhat lim
x→0
f (x)
x
= 1, it isnecessarilytruethat lim
x→0
f (x) = 0. For notethat lim
x→0
x = 0, whence
lim
x→0
f (x) = lim
x→0
x
f (x)
x
=
_
lim
x→0
x
__
lim
x→0
f (x)
x
_
= 0· 1= 0.
Provethat if lim
x→c
f (x) = L = 0and lim
x→c
g(x) = 0, thenthelimit lim
x→c
f (x)
g(x)
doesnot exist.
39. Supposethat lim
h→0
g(h) = L.
(a) Explainwhy lim
h→0
g(ah) = L for anyconstant a = 0.
(b) If weassumeinsteadthat lim
h→1
g(h) = L, isit still necessarilytruethat lim
h→1
g(ah) = L?
(c) Illustrate(a) and(b) withthefunctionf (x) = x
2
.
solution
(a) Ash → 0, ah → 0aswell; hence, if wemakethechangeof variablew = ah, then
lim
h→0
g(ah) = lim
w→0
g(w) = L.
(b) No. Ash → 1, ah → a, soweshouldnot expect lim
h→1
g(ah) = lim
h→1
g(h).
(c) Let g(x) = x
2
. Then
lim
h→0
g(h) = 0 and lim
h→0
g(ah) = lim
h→0
(ah)
2
= 0.
Ontheother hand,
lim
h→1
g(h) = 1 while lim
h→1
g(ah) = lim
h→1
(ah)
2
= a
2
,
whichisequal tothepreviouslimit if andonlyif a = ±1.
Assumethat L(a) = lim
x→0
a
x
−1
x
existsfor all a > 0. Assumealsothat lim
x→0
a
x
= 1.
(a) ProvethatL(ab) = L(a) +L(b) for a, b > 0. Hint: (ab)
x
−1= a
x
(b
x
−1) +(a
x
−1). ThisshowsthatL(a)
“behaves” likealogarithm. Wewill seethat L(a) = lna inSection7.3.
(b) Verifynumericallythat L(12) = L(3) +L(4).
2.4 Limits and Continuity
Preliminary Questions
1. Whichpropertyof f (x) = x
3
allowsustoconcludethat lim
x→2
x
3
= 8?
solution Wecanconcludethat lim
x→2
x
3
= 8becausethefunctionx
3
iscontinuousat x = 2.
2. What canbesaidabout f (3) if f iscontinuousand lim
x→3
f (x) =
1
2
?
solution If f iscontinuousandlim
x→3
f (x) =
1
2
, thenf (3) =
1
2
.
3. Supposethat f (x) < 0if x ispositiveandf (x) > 1if x isnegative. Canf becontinuousat x = 0?
solution Sincef (x) < 0whenxispositiveandf (x) > 1whenxisnegative, it followsthat
lim
x→0+
f (x) ≤ 0 and lim
x→0−
f (x) ≥ 1.
Thus, lim
x→0
f (x) doesnot exist, sof cannot becontinuousat x = 0.
June 7, 2011 LTSV SSM Second Pass
50 C HA P T E R 2 LIMITS
4. Is it possible to determine f (7) if f (x) = 3 for all x < 7 and f is right-continuous at x = 7? What if f is
left-continuous?
solution No. To determinef (7), weneedto combineeither knowledgeof thevalues of f (x) for x < 7withleft-
continuityor knowledgeof thevaluesof f (x) for x > 7withright-continuity.
5. Arethefollowingtrueor false? If false, stateacorrect version.
(a) f (x) iscontinuousat x = a if theleft- andright-handlimitsof f (x) asx → a exist andareequal.
(b) f (x) iscontinuousat x = a if theleft- andright-handlimitsof f (x) asx → a exist andequal f (a).
(c) If theleft- andright-handlimitsof f (x) asx → a exist, thenf hasaremovablediscontinuityat x = a.
(d) If f (x) andg(x) arecontinuousat x = a, thenf (x) +g(x) iscontinuousat x = a.
(e) If f (x) andg(x) arecontinuousat x = a, thenf (x)/g(x) iscontinuousat x = a.
solution
(a) False. Thecorrect statement is“f (x) iscontinuousat x = a if theleft- andright-handlimitsof f (x) asx → a exist
andequal f (a).”
(b) True.
(c) False. Thecorrect statement is“If theleft- andright-handlimitsof f (x) asx → a areequal but not equal tof (a),
thenf hasaremovablediscontinuityat x = a.”
(d) True.
(e) False. Thecorrectstatementis“If f (x) andg(x) arecontinuousatx = a andg(a) = 0, thenf (x)/g(x) iscontinuous
at x = a.”
Exercises
1. ReferringtoFigure14, statewhether f (x) isleft- or right-continuous(or neither) ateachpointof discontinuity. Does
f (x) haveanyremovablediscontinuities?
1 2 3 4 5 6
x
5
4
3
2
1
y
FIGURE 14 Graphof y = f (x)
solution
• Thefunctionf isdiscontinuousat x = 1; it isright-continuousthere.
• Thefunctionf isdiscontinuousat x = 3; it isneither left-continuousnor right-continuousthere.
• Thefunctionf isdiscontinuousat x = 5; it isleft-continuousthere.
However, thesediscontinuitiesarenot removable.
Exercises2–4refer tothefunctiong(x) inFigure15.
1 2 3 4 5 6
x
5
4
3
2
1
y
FIGURE 15 Graphof y = g(x)
Statewhether g(x) isleft- or right-continuous(or neither) at eachof itspointsof discontinuity.
3. At whichpoint c doesg(x) havearemovablediscontinuity? Howshouldg(c) beredeﬁnedtomakeg continuousat
x = c?
solution Becauselim
x→3
g(x) exists, thefunctiong has aremovablediscontinuity at x = 3. Assigningg(3) = 4
makesg continuousat x = 3.
Findthepointc
1
atwhichg(x) hasajumpdiscontinuitybutisleft-continuous. Howshouldg(c
1
) beredeﬁnedto
makeg right-continuousat x = c
1
?
June 7, 2011 LTSV SSM Second Pass
S E C T I ON 2.4 Limits and Continuity 51
5. InFigure16, determinetheone-sidedlimitsatthepointsof discontinuity. Whichdiscontinuityisremovableandhow
shouldf beredeﬁnedtomakeit continuousat thispoint?
4 2 −2
6
x
y
FIGURE 16
solution Thefunctionf isdiscontinuousat x = 0, at which lim
x→0−
f (x) = ∞and lim
x→0+
f (x) = 2. Thefunctionf
is also discontinuous at x = 2, at which lim
x→2−
f (x) = 6and lim
x→2+
f (x) = 6. Becausethetwo one-sidedlimits exist
andareequal at x = 2, thediscontinuityat x = 2isremovable. Assigningf (2) = 6makesf continuousat x = 2.
Supposethat f (x) = 2for x < 3andf (x) = −4for x > 3.
(a) What isf (3) if f isleft-continuousat x = 3?
(b) What isf (3) if f isright-continuousat x = 3?
InExercises7–16, usetheLawsof ContinuityandTheorems2and3toshowthat thefunctioniscontinuous.
7. f (x) = x +sinx
solution Sincex andsinx arecontinuous, soisx +sinx byContinuityLaw(i).
f (x) = x sinx
9. f (x) = 3x +4sinx
solution Sincex andsinx arecontinuous, soare3x and4sinx byContinuityLaw(ii).Thus3x +4sinx iscontinuous
byContinuityLaw(i).
f (x) = 3x
3
+8x
2
−20x
11. f (x) =
1
x
2
+1
solution
• Sincex iscontinuous, soisx
2
byContinuityLaw(iii).
• Recall that constant functions, suchas1, arecontinuous. Thusx
2
+1iscontinuous.
• Finally,
1
x
2
+1
iscontinuousbyContinuityLaw(iv) becausex
2
+1isnever 0.
f (x) =
x
2
−cosx
3+cosx
13. f (x) = cos(x
2
)
solution The function f (x) is a composite of two continuous functions: cosx and x
2
, so f (x) is continuous by
Theorem5, whichstatesthat acompositeof continuousfunctionsiscontinuous.
f (x) = sin(4
x
)
15. f (x) = 2
x
cos3x
solution 2
x
andcos3x arecontinuous, so2
x
cos3x iscontinuousbyContinuityLaw(iii).
f (x) = tan
_
1
x
2
+1
_
InExercises 17–34, determinethepoints of discontinuity. Statethetypeof discontinuity(removable, jump, inﬁnite, or
noneof these) andwhether thefunctionisleft- or right-continuous.
17. f (x) =
1
x
solution Thefunction1/x isdiscontinuousatx = 0, atwhichthereisaninﬁnitediscontinuity. Thefunctionisneither
left- nor right-continuousat x = 0.
f (x) = |x| 19. f (x) =
x −2
|x −1|
solution Thefunction
x −2
|x −1|
isdiscontinuousat x = 1, at whichthereisaninﬁnitediscontinuity. Thefunctionis
neither left- nor right-continuousat x = 1.
f (x) = [x] 21. f (x) =
_
1
2
x
_
solution Thefunction
_
1
2
x
_
isdiscontinuousat evenintegers, at whichtherearejumpdiscontinuities. Because
lim
x→2n+
_
1
2
x
_
= n
but
lim
x→2n−
_
1
2
x
_
= n −1,
it followsthat thisfunctionisright-continuousat theevenintegersbut not left-continuous.
June 7, 2011 LTSV SSM Second Pass
52 C HA P T E R 2 LIMITS
g(t ) =
1
t
2
−1
23. f (x) =
x +1
4x −2
solution Thefunction f (x) =
x +1
4x −2
is discontinuous at x =
1
2
, at which thereis an inﬁnitediscontinuity. The
functionisneither left- nor right-continuousat x =
1
2
.
h(z) =
1−2z
z
2
−z −6
25. f (x) = 3x
2/3
−9x
3
solution Thefunctionf (x) = 3x
2/3
−9x
3
isdeﬁnedandcontinuousfor all x.
g(t ) = 3t
−2/3
−9t
3
27. f (x) =

x −2
|x −2|
x = 2
−1 x = 2
solution For x > 2, f (x) =
x −2
(x −2)
= 1. For x < 2, f (x) =
(x −2)
(2−x)
= −1. Thefunctionhasajumpdiscontinuity
at x = 2. Because
lim
x→2−
f (x) = −1= f (2)
but
lim
x→2+
f (x) = 1= f (2),
it followsthat thisfunctionisleft-continuousat x = 2but not right-continuous.
f (x) =
_
cos
1
x
x = 0
1 x = 0
29. g(t ) = tan2t
solution Thefunctiong(t ) = tan2t =
sin2t
cos2t
isdiscontinuouswhenever cos2t = 0; i.e., whenever
2t =
(2n +1)π
2
or t =
(2n +1)π
4
,
wheren is aninteger. At every suchvalueof t thereis aninﬁnitediscontinuity. Thefunctionis neither left- nor right-
continuousat anyof thesepointsof discontinuity.
f (x) = csc(x
2
)
31. f (x) = tan(sinx)
solution Thefunctionf (x) = tan(sinx) iscontinuouseverywhere. Reason: sinx iscontinuouseverywhereandtanu
iscontinuouson
_

π
2
,
π
2
_
—andinparticular on−1≤ u = sinx ≤ 1. Continuityof tan(sinx) followsbythecontinuity
of compositefunctions.
f (x) = cos(π[x]) 33. f (x) =
1
2
x
−2
−x
solution Thefunctionf (x) =
1
2
x
−2
−x
isdiscontinuousat x = 0, at whichthereisaninﬁnitediscontinuity. The
functionisneither left- nor right-continuousat x = 0.
f (x) = 2
_
x
2
_
−4
_
x
4
_
InExercises35–48, determinethedomainof thefunctionandprovethat it iscontinuousonitsdomainusingtheLawsof
Continuityandthefactsquotedinthissection.
35. f (x) = 2sinx +3cosx
solution Thedomainof 2sinx +3cosx isall real numbers. Bothsinx andcosx arecontinuousonthisdomain, so
2sinx +3cosx iscontinuousbyContinuityLaws(i) and(ii).
f (x) =
_
x
2
+9
37. f (x) =

x sinx
solution Thisfunctionisdeﬁnedaslongasx ≥ 0. Since

x andsinx arecontinuous, sois

x sinx byContinuity
Law(iii).
f (x) =
x
2
x +x
1/4
39. f (x) = x
2/3
2
x
solution Thedomainof x
2/3
2
x
isall real numbersasthedenominator of therational exponentisodd. Bothx
2/3
and
2
x
arecontinuousonthisdomain, sox
2/3
2
x
iscontinuousbyContinuityLaw(iii).
f (x) = x
1/3
+x
3/4
41. f (x) = x
−4/3
solution This functionis deﬁnedfor all x = 0. Becausethefunctionx
4/3
is continuous andnot equal to zero for
x = 0, it followsthat
x
−4/3
=
1
x
4/3
iscontinuousfor x = 0byContinuityLaw(iv).
June 7, 2011 LTSV SSM Second Pass
S E C T I ON 2.4 Limits and Continuity 53
f (x) = cos
3
x
43. f (x) = tan
2
x
solution Thedomainof tan
2
x isall x = ±(2n −1)π/2wheren isapositiveinteger. Becausetanx iscontinuouson
thisdomain, it followsfromContinuityLaw(iii) that tan
2
x isalsocontinuousonthisdomain.
f (x) = cos(2
x
)
45. f (x) = (x
4
+1)
3/2
solution Thedomainof (x
4
+ 1)
3/2
isall real numbersasx
4
+ 1 > 0for all x. Becausex
3/2
andthepolynomial
x
4
+1arebothcontinuous, soisthecompositefunction(x
4
+1)
3/2
.
f (x) = 3
−x
2
47. f (x) =
cos(x
2
)
x
2
−1
solution Thedomain for this function is all x = ±1. Becausethefunctions cosx and x
2
arecontinuous on this
domain, soisthecompositefunctioncos(x
2
). Finally, becausethepolynomial x
2
−1iscontinuousandnotequal tozero
for x = ±1, thefunction
cos(x
2
)
x
2
−1
iscontinuousbyContinuityLaw(iv).
f (x) = 9
tanx
49. Showthat thefunction
f (x) =

x
2
+3 for x < 1
10−x for 1≤ x ≤ 2
6x −x
2
for x > 2
is continuous for x = 1, 2. Then computetheright- and left-hand limits at x = 1, 2, and determinewhether f (x) is
left-continuous, right-continuous, or continuousat thesepoints(Figure17).
6 2 1
9
y = 10 − x
y = 6x − x
2
y = x
2
+ 3
x
y
FIGURE 17
solution Let’sstart withx = 1, 2.
• Becausex iscontinuous, soisx
2
byContinuityLaw(iii). Theconstant function3isalsocontinuous, sox
2
+3is
continuousbyContinuityLaw(i). Therefore, f (x) iscontinuousfor x < 1.
• Becausex andtheconstant function10arecontinuous, thefunction10− x is continuous by Continuity Law(i).
Therefore, f (x) iscontinuousfor 1< x < 2.
• Becausex is continuous, x
2
is continuous by Continuity Law(iii) and6x is continuous by Continuity Law(ii).
Therefore, 6x −x
2
iscontinuousbyContinuityLaw(i), sof (x) iscontinuousfor x > 2.
At x = 1, f (x) hasajumpdiscontinuitybecausetheone-sidedlimitsexist but arenot equal:
lim
x→1−
f (x) = lim
x→1−
(x
2
+3) = 4, lim
x→1+
f (x) = lim
x→1+
(10−x) = 9.
Furthermore, theright-handlimit equalsthefunctionvaluef (1) = 9, sof (x) isright-continuousat x = 1. At x = 2,
lim
x→2−
f (x) = lim
x→2−
(10−x) = 8, lim
x→2+
f (x) = lim
x→2+
(6x −x
2
) = 8.
Theleft- andright-handlimitsexist andareequal tof (2), sof (x) iscontinuousat x = 2.
SawtoothFunction Draw the graph of f (x) = x − [x]. At which points is f discontinuous? Is it left- or
right-continuousat thosepoints?
InExercises51–54, sketchthegraphof f (x). At eachpoint of discontinuity, statewhether f isleft- or right-continuous.
51. f (x) =
_
x
2
for x ≤ 1
2−x for x > 1
solution
−1
1
−1
x
y
1 2 3
Thefunctionf iscontinuouseverywhere.
June 7, 2011 LTSV SSM Second Pass
54 C HA P T E R 2 LIMITS
f (x) =

x +1 for x < 1
1
x
for x ≥ 1
53. f (x) =

x
2
−3x +2
|x −2|
x = 2
0 x = 2
solution
1
−1
−2 4 6
2
3
4
5
y
x
Thefunctionf isneither left- nor right-continuousat x = 2.
f (x) =

x
3
+1 for −∞ < x ≤ 0
−x +1 for 0< x < 2
−x
2
+10x −15 for x ≥ 2
55. Showthat thefunction
f (x) =

x
2
−16
x −4
x = 4
10 x = 4
hasaremovablediscontinuityat x = 4.
solution Toshowthat f (x) hasaremovablediscontinuityat x = 4, wemust establishthat
lim
x→4
f (x)
existsbut doesnot equal f (4). Now,
lim
x→4
x
2
−16
x −4
= lim
x→4
(x +4) = 8= 10= f (4);
thus, f (x) hasaremovablediscontinuityat x = 4. Toremovethediscontinuity, wemust redeﬁnef (4) = 8.
Deﬁnef (x) = x sin
1
x
+2forx = 0. Plotf (x). Howshouldf (0) bedeﬁnedsothatf iscontinuousatx = 0?
InExercises57–59, ﬁndthevalueof theconstant (a, b, or c) that makesthefunctioncontinuous.
57. f (x) =
_
x
2
−c for x < 5
4x +2c for x ≥ 5
solution Asx → 5−, wehavex
2
− c → 25− c = L. Asx → 5+, wehave4x + 2c → 20+ 2c = R. Matchthe
limits: L = R or 25−c = 20+2c impliesc =
5
3
.
f (x) =
_
2x +9x
−1
for x ≤ 3
−4x +c for x > 3
59. f (x) =

x
−1
for x < −1
ax +b for −1≤ x ≤
1
2
x
−1
for x >
1
2
solution Asx → −1−, x
−1
→ −1whileasx → −1+, ax + b → b − a. For f tobecontinuousat x = −1, we
must thereforehaveb − a = −1. Now, as x →
1
2
−, ax + b →
1
2
a + b whileas x →
1
2
+, x
−1
→ 2. For f to be
continuous at x =
1
2
, wemust thereforehave
1
2
a + b = 2. Solvingthesetwo equations for a andb yields a = 2and
b = 1.
Deﬁne
g(x) =

x +3 for x < −1
cx for −1≤ x ≤ 2
x +2 for x > 2
Findavalueof c suchthat g(x) is
(a) left-continuous (b) right-continuous
Ineachcase, sketchthegraphof g(x).
61. Deﬁneg(t ) = 2
1/(t −1)
for t = 0. Answer thefollowingquestions, usingaplot if necessary.
(a) Cang(1) bedeﬁnedsothat g(t ) iscontinuousat t = 1?
(b) Howshouldg(1) bedeﬁnedsothat g(t ) isleft-continuousat t = 1?
solution
(a) Fromthegraphof g(t ) shownbelow, weseethat g approaches 0as t → 1fromtheleft andbecomes inﬁniteas
t → 1fromtheright. Therefore, g(1) cannot bedeﬁnedsothat g iscontinuousat t = 1.
1
0.5
0
1.0 1.5 2.0
2
3
4
5
y
t
June 7, 2011 LTSV SSM Second Pass
S E C T I ON 2.4 Limits and Continuity 55
(b) Tomakeg(t ) left-continuousat t = 1, weshoulddeﬁne
g(1) = lim
t →1

2
1/(t −1)
= 0
Each of the following statements is false. For each statement, sketch the graph of a function that provides a
counterexample.
(a) If lim
x→a
f (x) exists, thenf (x) iscontinuousat x = a.
(b) If f (x) hasajumpdiscontinuityat x = a, thenf (a) isequal toeither lim
x→a−
f (x) or lim
x→a+
f (x).
InExercises63–66, drawthegraphof afunctionon[0, 5] withthegivenproperties.
63. f (x) isnot continuousat x = 1, but lim
x→1+
f (x) and lim
x→1−
f (x) exist andareequal.
solution
5 4 3 2 1
1
2
3
4
y
x
f (x) isleft-continuousbut not continuousat x = 2andright-continuousbut not continuousat x = 3.
65. f (x) hasaremovablediscontinuityat x = 1, ajumpdiscontinuityat x = 2, and
lim
x→3−
f (x) = −∞, lim
x→3+
f (x) = 2
solution
5 4 3 2 1
1
2
3
4
y
x
f (x) is right- but not left-continuous at x = 1, left- but not right-continuous at x = 2, and neither left- nor
right-continuousat x = 3.
InExercises67–80, evaluateusingsubstitution.
67. lim
x→−1
(2x
3
−4)
solution lim
x→−1
(2x
3
−4) = 2(−1)
3
−4= −6.
lim
x→2
(5x −12x
−2
)
69. lim
x→3
x +2
x
2
+2x
solution lim
x→3
x +2
x
2
+2x
=
3+2
3
2
+2· 3
=
5
15
=
1
3
lim
x→π
sin
_
x
2
−π
_ 71. lim
x→
π
4
tan(3x)
solution lim
x→
π
4
tan(3x) = tan(3·
π
4
) = tan(

4
) = −1
lim
x→π
1
cosx
73. lim
x→4
x
−5/2
solution lim
x→4
x
−5/2
= 4
−5/2
=
1
32
.
lim
x→2
_
x
3
+4x
75. lim
x→−1
(1−8x
3
)
3/2
solution lim
x→−1
(1−8x
3
)
3/2
= (1−8(−1)
3
)
3/2
= 27.
lim
x→2
_
7x +2
4−x
_
2/3
77. lim
x→3
10
x
2
−2x
solution lim
x→3
10
x
2
−2x
= 10
3
2
−2(3)
= 1000.
June 7, 2011 LTSV SSM Second Pass
56 C HA P T E R 2 LIMITS
lim
x→−
π
2
3
sinx
79. lim
x→
π
3
sin
2
(π sin
2
x)
solution lim
x→
π
3
sin
2
(π sin
2
x) = sin
2
_
π lim
x→
π
3
sin
2
x
_
= sin
2
_
π sin
2
_
π
3
_
_
= sin
2
_

4
_
=
1
2
lim
x→1
tan
_
2
x−1
_ 81. Supposethat f (x) andg(x) arediscontinuousat x = c. Doesit followthat f (x) +g(x) isdiscontinuousat x = c?
If not, giveacounterexample. Doesthiscontradict Theorem1(i)?
solution Evenif f (x) andg(x) arediscontinuousatx = c, itisnot necessarilytruethatf (x) +g(x) isdiscontinuous
atx = c. Forexample, supposef (x) = −x
−1
andg(x) = x
−1
. Bothf (x) andg(x) arediscontinuousatx = 0; however,
thefunctionf (x) +g(x) = 0, whichiscontinuouseverywhere, includingx = 0. ThisdoesnotcontradictTheorem1(i),
whichdealsonlywithcontinuousfunctions.
Provethat f (x) = |x| iscontinuousfor all x. Hint: Toprovecontinuityat x = 0, consider theone-sidedlimits.
83. Usetheresult of Exercise82toprovethat if g(x) iscontinuous, thenf (x) = |g(x)| isalsocontinuous.
solution Recall thatthecompositionof twocontinuousfunctionsiscontinuous. Now, f (x) = |g(x)| isacomposition
of thecontinuousfunctionsg(x) and|x|, soisalsocontinuous.
Whichof thefollowingquantitieswouldberepresentedby continuousfunctionsof timeandwhichwouldhave
oneor morediscontinuities?
(a) Velocityof anairplaneduringaﬂight
(b) Temperatureinaroomunder ordinaryconditions
(c) Valueof abankaccount withinterest paidyearly
(d) Thesalaryof ateacher
(e) Thepopulationof theworld
85. In2009, thefederal incometaxT (x) onincomeof x dollars(upto\$82,250) wasdeterminedbytheformula
T (x) =

0.10x for 0≤ x < 8350
0.15x −417.50 for 8350≤ x < 33,950
0.25x −3812.50 for 33,950≤ x < 82,250
Sketchthegraphof T (x). DoesT (x) haveanydiscontinuities? Explainwhy, if T (x) hadajumpdiscontinuity, it might
beadvantageousinsomesituationstoearnlessmoney.
solution T (x), theamount of federal incometaxowedonanincomeof x dollarsin2009, might beadiscontinuous
functiondependinguponhowthetaxtablesareconstructed(asdeterminedbythatyear’sregulations). Hereisagraphof
T (x) for that particular year.
20,000
5000
10,000
15,000
40,000 60,000 80,000
x
y
If T (x) hadajumpdiscontinuity(sayat x = c), it might beadvantageoustoearnslightlylessincomethanc (sayc −)
andbetaxedat alower ratethantoearnc or moreandbetaxedat ahigher rate. Your net earningsmayactuallybemore
intheformer casethaninthelatter one.
Further Insights and Challenges
If f (x) hasaremovablediscontinuityat x = c, thenit ispossibletoredeﬁnef (c) sothat f (x) iscontinuous
at x = c. Canthisbedoneinmorethanoneway?
87. Giveanexampleof functionsf (x) andg(x) suchthat f (g(x)) iscontinuousbut g(x) hasat least onediscontinuity.
solution Answersmayvary. Thesimplestexamplesarethefunctionsf (g(x)) wheref (x) = C isaconstantfunction,
andg(x) isdeﬁnedfor all x. Inthesecases, f (g(x)) = C. For example, if f (x) = 3andg(x) = [x], g isdiscontinuous
at all integer valuesx = n, but f (g(x)) = 3iscontinuous.
ContinuousatOnlyOnePoint Showthat thefollowingfunctioniscontinuousonlyat x = 0:
f (x) =
_
x for x rational
−x for x irrational
89. Showthat f (x) isadiscontinuousfunctionfor all x wheref (x) isdeﬁnedasfollows:
f (x) =
_
1 for x rational
−1 for x irrational
Showthat f (x)
2
iscontinuousfor all x.
solution lim
x→c
f (x) doesnotexistfor anyc. If c isirrational, thenthereisalwaysarational number r arbitrarilyclose
toc sothat |f (c) −f (r)| = 2. If, ontheother hand, c isrational, thereisalwaysanirrational number z arbitrarilyclose
toc sothat |f (c) −f (z)| = 2.
Ontheother hand, f (x)
2
isaconstant functionthat alwayshasvalue1, whichisobviouslycontinuous.
June 7, 2011 LTSV SSM Second Pass
S E C T I ON 2.5 Evaluating Limits Algebraically 57
2.5 Evaluating Limits Algebraically
Preliminary Questions
1. Whichof thefollowingisindeterminateat x = 1?
x
2
+1
x −1
,
x
2
−1
x +2
,
x
2
−1

x +3−2
,
x
2
+1

x +3−2
solution At x = 1,
x
2
−1

x+3−2
isof theform
0
0
; hence, thisfunctionisindeterminate. Noneof theremainingfunctions
is indeterminateat x = 1:
x
2
+1
x−1
and
x
2
+1

x+3−2
areundeﬁnedbecausethedenominator is zerobut thenumerator is not,
while
x
2
−1
x+2
isequal to0.
2. Givecounterexamplestoshowthat thesestatementsarefalse:
(a) If f (c) isindeterminate, thentheright- andleft-handlimitsasx → c arenot equal.
(b) If lim
x→c
f (x) exists, thenf (c) isnot indeterminate.
(c) If f (x) isundeﬁnedat x = c, thenf (x) hasanindeterminateformat x = c.
solution
(a) Let f (x) =
x
2
−1
x−1
. At x = 1, f isindeterminateof theform
0
0
but
lim
x→1−
x
2
−1
x −1
= lim
x→1−
(x +1) = 2= lim
x→1+
(x +1) = lim
x→1+
x
2
−1
x −1
.
(b) Again, let f (x) =
x
2
−1
x−1
. Then
lim
x→1
f (x) = lim
x→1
x
2
−1
x −1
= lim
x→1
(x +1) = 2
but f (1) isindeterminateof theform
0
0
.
(c) Let f (x) =
1
x
. Thenf isundeﬁnedat x = 0but doesnot haveanindeterminateformat x = 0.
3. Themethodfor evaluatinglimitsdiscussedinthissectionissometimescalled“simplifyandplugin.” Explainhowit
actuallyreliesonthepropertyof continuity.
solution If f is continuous at x = c, then, by deﬁnition, lim
x→c
f (x) = f (c); in other words, the limit of a
continuous function at x = c is thevalueof thefunction at x = c. The“simplify and plug-in" strategy is based on
simplifyingafunctionwhichisindeterminatetoacontinuousfunction. Oncethesimpliﬁcationhasbeenmade, thelimit
of theremainingcontinuousfunctionisobtainedbyevaluation.
Exercises
InExercises1–4, showthat thelimit leadstoanindeterminateform. Thencarryout thetwo-stepprocedure: Transform
thefunctionalgebraicallyandevaluateusingcontinuity.
1. lim
x→6
x
2
−36
x −6
solution Whenwesubstitutex = 6into
x
2
−36
x−6
, weobtaintheindeterminateform
0
0
. Uponfactoringthenumerator
andsimplifying, weﬁnd
lim
x→6
x
2
−36
x −6
= lim
x→6
(x −6)(x +6)
x −6
= lim
x→6
(x +6) = 12.
lim
h→3
9−h
2
h −3
3. lim
x→−1
x
2
+2x +1
x +1
solution Whenwesubstitutex = −1into
x
2
+2x+1
x+1
,weobtaintheindeterminateform
0
0
.Uponfactoringthenumerator
andsimplifying, weﬁnd
lim
x→−1
x
2
+2x +1
x +1
= lim
x→−1
(x +1)
2
x +1
= lim
x→−1
(x +1) = 0.
June 7, 2011 LTSV SSM Second Pass
58 C HA P T E R 2 LIMITS
lim
t →9
2t −18
5t −45
InExercises5–34, evaluatethelimit, if it exists. If not, determinewhether theone-sidedlimitsexist (ﬁniteor inﬁnite).
5. lim
x→7
x −7
x
2
−49
solution lim
x→7
x −7
x
2
−49
= lim
x→7
x −7
(x −7)(x +7)
= lim
x→7
1
x +7
=
1
14
.
lim
x→8
x
2
−64
x −9
7. lim
x→−2
x
2
+3x +2
x +2
solution lim
x→−2
x
2
+3x +2
x +2
= lim
x→−2
(x +1)(x +2)
x +2
= lim
x→−2
(x +1) = −1.
lim
x→8
x
3
−64x
x −8
9. lim
x→5
2x
2
−9x −5
x
2
−25
solution lim
x→5
2x
2
−9x −5
x
2
−25
= lim
x→5
(x −5)(2x +1)
(x −5)(x +5)
= lim
x→5
2x +1
x +5
=
11
10
.
lim
h→0
(1+h)
3
−1
h
11. lim
x→−
1
2
2x +1
2x
2
+3x +1
solution lim
x→−
1
2
2x +1
2x
2
+3x +1
= lim
x→−
1
2
2x +1
(2x +1)(x +1)
= lim
x→−
1
2
1
x +1
= 2.
lim
x→3
x
2
−x
x
2
−9
13. lim
x→2
3x
2
−4x −4
2x
2
−8
solution lim
x→2
3x
2
−4x −4
2x
2
−8
= lim
x→2
(3x +2)(x −2)
2(x −2)(x +2)
= lim
x→2
3x +2
2(x +2)
=
8
8
= 1.
lim
h→0
(3+h)
3
−27
h
15. lim
t →0
4
2t
−1
4
t
−1
solution lim
t t o0
4
2t
−1
4
t
−1
= lim
t t o0
(4
t
−1)(4
t
+1)
4
t
−1
= lim
t →0
(4
t
+1) = 2.
lim
h→4
(h +2)
2
−9h
h −4
17. lim
x→16

x −4
x −16
solution lim
x→16

x −4
x −16
= lim
x→16

x −4
_√
x +4
_ _√
x −4
_ = lim
x→16
1

x +4
=
1
8
.
lim
t →−2
2t +4
12−3t
2
19. lim
y→3
y
2
+y −12
y
3
−10y +3
solution lim
y→3
y
2
+y −12
y
3
−10y +3
= lim
y→3
(y −3)(y +4)
(y −3)(y
2
+3y −1)
= lim
y→3
(y +4)
(y
2
+3y −1)
=
7
17
.
lim
h→0
1
(h +2)
2

1
4
h
21. lim
h→0

2+h −2
h
solution lim
h→0

h +2−2
h
doesnot exist.
• Ash → 0+, wehave

h +2−2
h
=
_√
h +2−2
_
(

h +2+2)
h(

h +2+2)
=
h −2
h(

h +2+2)
→ −∞.
• Ash → 0−, wehave

h +2−2
h
=
_√
h +2−2
_
(

h +2+2)
h(

h +2+2)
=
h −2
h(

h +2+2)
→ ∞.
lim
x→8

x −4−2
x −8
23. lim
x→4
x −4

x −

8−x
solution
lim
x→4
x −4

x −

8−x
= lim
x→4
(x −4)(

x +

8−x)
(

x −

8−x)(

x +

8−x)
= lim
x→4
(x −4)(

x +

8−x)
x −(8−x)
= lim
x→4
(x −4)(

x +

8−x)
2x −8
= lim
x→4
(x −4)(

x +

8−x)
2(x −4)
= lim
x→4
(

x +

8−x)
2
=

4+

4
2
= 2.
June 7, 2011 LTSV SSM Second Pass
S E C T I ON 2.5 Evaluating Limits Algebraically 59
lim
x→4

5−x −1
2−

x
25. lim
x→4
_
1

x −2

4
x −4
_
solution lim
x→4
_
1

x −2

4
x −4
_
= lim
x→4

x +2−4
_√
x −2
_ _√
x +2
_ = lim
x→4

x −2
_√
x −2
_ _√
x +2
_ =
1
4
.
lim
x→0+
_
1

x

1
_
x
2
+x
_
27. lim
x→0
cotx
cscx
solution lim
x→0
cotx
cscx
= lim
x→0
cosx
sinx
· sinx = cos0= 1.
lim
θ→
π
2
cotθ
cscθ
29. lim
t →2
2
2t
+2
t
−20
2
t
−4
solution lim
t →2
2
2t
+2
t
−20
2
t
−4
= lim
t →2
(2
t
+5)(2
t
−4)
2
t
−4
= lim
t →2
(2
t
+5) = 9.
lim
x→1
_
1
1−x

2
1−x
2
_
31. lim
x→
π
4
sinx −cosx
tanx −1
solution lim
x→
π
4
sinx −cosx
tanx −1
·
cosx
cosx
= lim
x→
π
4
(sinx −cosx) cosx
sinx −cosx
= cos
π
4
=

2
2
.
lim
θ→
π
2
_
secθ −tanθ
_
33. lim
θ→
π
4
_
1
tanθ −1

2
tan
2
θ −1
_
solution lim
θ→
π
4
_
1
tanθ −1

2
tan
2
θ −1
_
= lim
θ→
π
4
(tanθ +1) −2
(tanθ +1)(tanθ −1)
= lim
θ→
π
4
1
tanθ +1
=
1
2
.
lim
x→
π
3
2cos
2
x +3cosx −2
2cosx −1
35. Useaplot of f (x) =
x −4

x −

8−x
toestimate lim
x→4
f (x) totwodecimal places. Comparewiththeanswer
obtainedalgebraicallyinExercise23.
solution Letf (x) =
x−4

x−

8−x
. Fromtheplotof f (x) shownbelow, weestimate lim
x→4
f (x) ≈ 2.00; totwodecimal
places, thismatchesthevalueof 2obtainedinExercise23.
1.996
1.997
1.998
1.999
2.000
2.001
1.995
3.6
y
x
3.8 4.0 4.2 4.4
Useaplotof f (x) =
1

x −2

4
x −4
toestimate lim
x→4
f (x) numerically. Comparewiththeanswer obtained
algebraicallyinExercise25.
InExercises37–42, evaluateusingtheidentity
a
3
−b
3
= (a −b)(a
2
+ab +b
2
)
37. lim
x→2
x
3
−8
x −2
solution lim
x→2
x
3
−8
x −2
= lim
x→2
(x −2)
_
x
2
+2x +4
_
x −2
= lim
x→2
_
x
2
+2x +4
_
= 12.
lim
x→3
x
3
−27
x
2
−9
39. lim
x→1
x
2
−5x +4
x
3
−1
solution lim
x→1
x
2
−5x +4
x
3
−1
= lim
x→1
(x −1)(x −4)
(x −1)
_
x
2
+x +1
_ = lim
x→1
x −4
x
2
+x +1
= −1.
lim
x→−2
x
3
+8
x
2
+6x +8
41. lim
x→1
x
4
−1
x
3
−1
solution
lim
x→1
x
4
−1
x
3
−1
= lim
x→1
(x
2
−1)(x
2
+1)
(x −1)(x
2
+x +1)
= lim
x→1
(x −1)(x +1)(x
2
+1)
(x −1)(x
2
+x +1)
= lim
x→1
(x +1)(x
2
+1)
(x
2
+x +1)
=
4
3
.
June 7, 2011 LTSV SSM Second Pass
60 C HA P T E R 2 LIMITS
lim
x→27
x −27
x
1/3
−3
43. Evaluate lim
h→0
4

1+h −1
h
. Hint: Set x =
4

1+h andrewriteasalimit asx → 1.
solution Let x =
4

1+h. Thenh = x
4
−1= (x −1)(x +1)(x
2
+1), x → 1ash → 0and
lim
h→0
4

1+h −1
h
= lim
x→1
x −1
(x −1)(x +1)(x
2
+1)
= lim
x→1
1
(x +1)(x
2
+1)
=
1
4
.
Evaluate lim
h→0
3

1+h −1
2

1+h −1
. Hint: Set x =
6

1+h andrewriteasalimit asx → 1.
InExercises45–54, evaluateintermsof theconstant a.
45. lim
x→0
(2a +x)
solution lim
x→0
(2a +x) = 2a.
lim
h→−2
(4ah +7a)
47. lim
t →−1
(4t −2at +3a)
solution lim
t →−1
(4t −2at +3a) = −4+5a.
lim
h→0
(3a +h)
2
−9a
2
h
49. lim
h→0
2(a +h)
2
−2a
2
h
solution lim
h→0
2(a +h)
2
−2a
2
h
= lim
h→0
4ha +2h
2
h
= lim
h→0
(4a +2h) = 4a.
lim
x→a
(x +a)
2
−4x
2
x −a
51. lim
x→a

x −

a
x −a
solution lim
x→a

x −

a
x −a
= lim
x→a

x −

a
_√
x −

a
_ _√
x +

a
_ = lim
x→a
1

x +

a
=
1
2

a
.
lim
h→0

a +2h −

a
h
53. lim
x→0
(x +a)
3
−a
3
x
solution lim
x→0
(x +a)
3
−a
3
x
= lim
x→0
x
3
+3x
2
a +3xa
2
+a
3
−a
3
x
= lim
x→0
(x
2
+3xa +3a
2
) = 3a
2
.
lim
h→a
1
h

1
a
h −a
Further Insights and Challenges
InExercises55–58, ﬁndall valuesof c suchthat thelimit exists.
55. lim
x→c
x
2
−5x −6
x −c
solution lim
x→c
x
2
−5x −6
x −c
will exist providedthat x − c is afactor of thenumerator. (Otherwisetherewill bean
inﬁnitediscontinuityat x = c.) Sincex
2
−5x −6= (x +1)(x −6), thisoccursfor c = −1andc = 6.
lim
x→1
x
2
+3x +c
x −1
57. lim
x→1
_
1
x −1

c
x
3
−1
_
solution Simplifying, weﬁnd
1
x −1

c
x
3
−1
=
x
2
+x +1−c
(x −1)(x
2
+x +1)
.
Inorder for thelimit toexist asx → 1, thenumerator must evaluateto0at x = 1. Thus, wemust have3−c = 0, which
impliesc = 3.
lim
x→0
1+cx
2

_
1+x
2
x
4
59. For whichsign± doesthefollowinglimit exist?
lim
x→0
_
1
x
±
1
x(x −1)
_
solution
• Thelimit lim
x→0
_
1
x
+
1
x(x −1)
_
= lim
x→0
(x −1) +1
x(x −1)
= lim
x→0
1
x −1
= −1.
• Thelimit lim
x→0
_
1
x

1
x(x −1)
_
doesnot exist.
– Asx → 0+, wehave
1
x

1
x(x −1)
=
(x −1) −1
x(x −1)
=
x −2
x(x −1)
→ ∞.
– Asx → 0−, wehave
1
x

1
x(x −1)
=
(x −1) −1
x(x −1)
=
x −2
x(x −1)
→ −∞.
June 7, 2011 LTSV SSM Second Pass
S E C T I ON 2.6 Trigonometric Limits 61
2.6 Trigonometric Limits
Preliminary Questions
1. Assumethat −x
4
≤ f (x) ≤ x
2
. What is lim
x→0
f (x)? Isthereenoughinformationtoevaluate lim
x→
1
2
f (x)? Explain.
solution Since lim
x→0
−x
4
= lim
x→0
x
2
= 0, the squeeze theoremguarantees that lim
x→0
f (x) = 0. Since
lim
x→
1
2
−x
4
= −
1
16
=
1
4
= lim
x→
1
2
x
2
, wedonot haveenoughinformationtodeterminelim
x→
1
2
f (x).
2. StatetheSqueezeTheoremcarefully.
solution Assumethat for x = c (insomeopeninterval containingc),
l(x) ≤ f (x) ≤ u(x)
andthat lim
x→c
l(x) = lim
x→c
u(x) = L. Then lim
x→c
f (x) existsand
lim
x→c
f (x) = L.
3. If youwant toevaluate lim
h→0
sin5h
3h
, it isagoodideatorewritethelimit intermsof thevariable(chooseone):
(a) θ = 5h (b) θ = 3h (c) θ =
5h
3
solution Tomatchthegivenlimit tothepatternof
lim
θ→0
sinθ
θ
,
it isbest tosubstitutefor theargument of thesinefunction; thus, rewritethelimit intermsof (a): θ = 5h.
Exercises
1. Statepreciselythehypothesisandconclusionsof theSqueezeTheoremfor thesituationinFigure6.
1 2
2
u(x)
l(x)
f(x)
x
y
FIGURE 6
solution For all x = 1ontheopeninterval (0, 2) containingx = 1, (x) ≤ f (x) ≤ u(x). Moreover,
lim
x→1
(x) = lim
x→1
u(x) = 2.
Therefore, bytheSqueezeTheorem,
lim
x→1
f (x) = 2.
InFigure7, isf (x) squeezedbyu(x) andl(x) at x = 3?At x = 2?
3. WhatdoestheSqueezeTheoremsayabout lim
x→7
f (x) if lim
x→7
l(x) = lim
x→7
u(x) = 6andf (x),u(x),andl(x) arerelated
asinFigure8?Theinequalityf (x) ≤ u(x) isnot satisﬁedfor all x. Doesthisaffect thevalidityof your conclusion?
7
6
x
u(x)
f(x)
l(x)
y
FIGURE 8
June 7, 2011 LTSV SSM Second Pass
62 C HA P T E R 2 LIMITS
solution TheSqueezeTheoremdoesnotrequirethattheinequalitiesl(x) ≤ f (x) ≤ u(x) holdfor all x, onlythatthe
inequalitiesholdonsomeopeninterval containingx = c. InFigure8, it isclear that l(x) ≤ f (x) ≤ u(x) onsomeopen
interval containingx = 7. Because lim
x→7
u(x) = lim
x→7
l(x) = 6, theSqueezeTheoremguaranteesthat lim
x→7
f (x) = 6.
Determine lim
x→0
f (x) assumingthat cosx ≤ f (x) ≤ 1.
5. Statewhether theinequalityprovidessufﬁcient informationtodetermine lim
x→1
f (x), andif so, ﬁndthelimit.
(a) 4x −5≤ f (x) ≤ x
2
(b) 2x −1≤ f (x) ≤ x
2
(c) 4x −x
2
≤ f (x) ≤ x
2
+2
solution
(a) Because lim
x→1
(4x −5) = −1= 1= lim
x→1
x
2
,thegiveninequalitydoesnotprovidesufﬁcientinformationtodetermine
lim
x→1
f (x).
(b) Because lim
x→1
(2x −1) = 1= lim
x→1
x
2
, it followsfromtheSqueezeTheoremthat lim
x→1
f (x) = 1.
(c) Because lim
x→1
(4x −x
2
) = 3= lim
x→1
(x
2
+2), it followsfromtheSqueezeTheoremthat lim
x→1
f (x) = 3.
Plot thegraphs of u(x) = 1+
¸
¸
x −
π
2
¸
¸
andl(x) = sinx onthesameset of axes. What canyousay about
lim
x→
π
2
f (x) if f (x) issqueezedbyl(x) andu(x) at x =
π
2
?
InExercises7–16, evaluateusingtheSqueezeTheorem.
7. lim
x→0
x
2
cos
1
x
solution Multiplyingtheinequality−1≤ cos
1
x
≤ 1, whichholdsfor all x = 0, byx
2
yields−x
2
≤ x
2
cos
1
x
≤ x
2
.
Because
lim
x→0
−x
2
= lim
x→0
x
2
= 0,
it followsbytheSqueezeTheoremthat
lim
x→0
x
2
cos
1
x
= 0.
lim
x→0
x sin
1
x
2
9. lim
x→1
(x −1) sin
π
x −1
solution Multiplyingtheinequality
¸
¸
¸sin
π
x−1
¸
¸
¸ ≤ 1, whichholds for x = 1, by |x − 1| yields
¸
¸
¸(x −1) sin
π
x−1
¸
¸
¸ ≤
|x −1| or −|x −1| ≤ (x −1) sin
π
x−1
≤ |x −1|. Because
lim
x→1
−|x −1| = lim
x→1
|x −1| = 0,
it followsbytheSqueezeTheoremthat
lim
x→1
(x −1) sin
π
x −1
= 0.
lim
x→3
(x
2
−9)
x −3
|x −3|
11. lim
t →0
(2
t
−1) cos
1
t
solution Multiplyingtheinequality
¸
¸
¸cos
1
t
¸
¸
¸ ≤ 1, whichholdsfor t = 0, by|2
t
−1| yields
¸
¸
¸(2
t
−1) cos
1
t
¸
¸
¸ ≤ |2
t
−1|
or −|2
t
−1| ≤ (2
t
−1) cos
1
t
≤ |2
t
−1|. Because
lim
t →0
−|2
t
−1| = lim
t →0
|2
t
−1| = 0,
it followsbytheSqueezeTheoremthat
lim
t →0
(2
t
−1) cos
1
t
= 0.
lim
x→0+

x 4
cos(π/x) 13. lim
t →2
(t
2
−4) cos
1
t −2
solution Multiplyingtheinequality
¸
¸
¸cos
1
t −2
¸
¸
¸ ≤ 1, whichholds for t = 2, by |t
2
− 4| yields
¸
¸
¸(t
2
−4) cos
1
t −2
¸
¸
¸ ≤
|t
2
−4| or −|t
2
−4| ≤ (t
2
−4) cos
1
t −2
≤ |t
2
−4|. Because
lim
t →2
−|t
2
−4| = lim
t →2
|t
2
−4| = 0,
June 7, 2011 LTSV SSM Second Pass
S E C T I ON 2.6 Trigonometric Limits 63
it followsbytheSqueezeTheoremthat
lim
t →2
(t
2
−4) cos
1
t −2
= 0.
lim
x→0
tanx cos
_
sin
1
x
_
15. lim
θ→
π
2
cosθ cos(tanθ)
solution Multiplyingtheinequality | cos(tanθ)| ≤ 1, whichholds for all θ near
π
2
but not equal to
π
2
, by | cosθ|
yields| cosθ cos(tanθ)| ≤ | cosθ| or −| cosθ| ≤ cosθ cos(tanθ) ≤ | cosθ|. Because
lim
θ→
π
2
−| cosθ| = lim
θ→
π
2
| cosθ| = 0,
it followsfromtheSqueezeTheoremthat
lim
θ→
π
2
cosθ cos(tanθ) = 0.
lim
t →0−
sin
2
_
1
t
_
3
1/t
InExercises17–26, evaluateusingTheorem2asnecessary.
17. lim
x→0
tanx
x
solution lim
x→0
tanx
x
= lim
x→0
sinx
x
1
cosx
= lim
x→0
sinx
x
· lim
x→0
1
cosx
= 1· 1= 1.
lim
x→0
sinx secx
x
19. lim
t →0
_
t
3
+9sint
t
solution lim
t →0
_
t
3
+9sint
t
= lim
t →0
_
t
3
+9· lim
t →0
sint
t
=

9· 1= 3.
lim
t →0
sin
2
t
t
21. lim
x→0
x
2
sin
2
x
solution lim
x→0
x
2
sin
2
x
= lim
x→0
1
sinx
x
sinx
x
= lim
x→0
1
sinx
x
· lim
x→0
1
sinx
x
=
1
1
·
1
1
= 1.
lim
t →
π
2
1−cost
t
23. lim
θ→0
secθ −1
θ
solution lim
θ→0
secθ −1
θ
= lim
θ→0
1−cosθ
θ cosθ
= lim
θ→0
1−cosθ
θ
· lim
θ→0
1
cosθ
= 0· 1= 0.
lim
θ→0
1−cosθ
sinθ
25. lim
t →
π
4
sint
t
solution
sint
t
iscontinuousat t =
π
4
. Hence, bysubstitution
lim
t →
π
4
sint
t
=

2
2
π
4
=
2

2
π
.
lim
t →0
cost −cos
2
t
t
27. Let L = lim
x→0
sin14x
x
.
(a) Show, bylettingθ = 14x, that L = lim
θ→0
14
sinθ
θ
.
(b) ComputeL.
solution
(a) Let θ = 14x. Thenx =
θ
14
andθ → 0asx → 0, so
L = lim
x→0
sin14x
x
= lim
θ→0
sinθ
(θ/14)
= lim
θ→0
14
sinθ
θ
.
(b) Basedonpart (a),
L = 14 lim
θ→0
·
sinθ
θ
= 14.
June 7, 2011 LTSV SSM Second Pass
64 C HA P T E R 2 LIMITS
Evaluate lim
h→0
sin9h
sin7h
. Hint:
sin9h
sin7h
=
_
9
7
__
sin9h
9h
__
7h
sin7h
_
.
InExercises29–48, evaluatethelimit.
29. lim
h→0
sin9h
h
solution lim
h→0
sin9h
h
= lim
h→0
9
sin9h
9h
= 9.
lim
h→0
sin4h
4h
31. lim
h→0
sinh
5h
solution lim
h→0
sinh
5h
= lim
h→0
1
5
sinh
h
=
1
5
.
lim
x→
π
6
x
sin3x
33. lim
θ→0
sin7θ
sin3θ
solution Wehave
sin7θ
sin3θ
=
7
3
_
sin7θ

__

sin3θ
_
Therefore,
lim
θ→0
sin7θ

=
7
3
_
lim
θ→0
sin7θ

__
lim
θ→0

sin3θ
_
=
7
3
(1)(1) =
7
3
lim
x→0
tan4x
9x
35. lim
x→0
x csc25x
solution Let h = 25x. Then
lim
x→0
x csc25x = lim
h→0
h
25
csch =
1
25
lim
h→0
h
sinh
=
1
25
.
lim
t →0
tan4t
t sect
37. lim
h→0
sin2h sin3h
h
2
solution
lim
h→0
sin2h sin3h
h
2
= lim
h→0
sin2h sin3h
h · h
= lim
h→0
sin2h
h
sin3h
h
= lim
h→0
2
sin2h
2h
3
sin3h
3h
= lim
h→0
2
sin2h
2h
lim
h→0
3
sin3h
3h
= 2· 3= 6.
lim
z→0
sin(z/3)
sinz
39. lim
θ→0
sin(−3θ)
sin(4θ)
solution lim
θ→0
sin(−3θ)
sin(4θ)
= lim
θ→0
−sin(3θ)

·
3
4
·

sin(4θ)
= −
3
4
.
lim
x→0
tan4x
tan9x
41. lim
t →0
csc8t
csc4t
solution lim
t →0
csc8t
csc4t
= lim
t →0
sin4t
sin8t
·
8t
4t
·
1
2
=
1
2
.
lim
x→0
sin5x sin2x
sin3x sin5x
43. lim
x→0
sin3x sin2x
x sin5x
solution lim
x→0
sin3x sin2x
x sin5x
= lim
x→0
_
3
sin3x
3x
·
2
5
(sin2x) / (2x)
(sin5x) / (5x)
_
=
6
5
.
lim
h→0
1−cos2h
h
45. lim
h→0
sin(2h)(1−cosh)
h
2
solution lim
h→0
sin(2h)(1−cosh)
h
2
= lim
h→0
sin(2h)
h
lim
h→0
1−cosh
h
= 1· 0= 0.
lim
t →0
1−cos2t
sin
2
3t
June 7, 2011 LTSV SSM Second Pass
S E C T I ON 2.6 Trigonometric Limits 65
47. lim
θ→0
cos2θ −cosθ
θ
solution
lim
θ→0
cos2θ −cosθ
θ
= lim
θ→0
(cos2θ −1) +(1−cosθ)
θ
= lim
θ→0
cos2θ −1
θ
+ lim
θ→0
1−cosθ
θ
= −2 lim
θ→0
1−cos2θ

+ lim
θ→0
1−cosθ
θ
= −2· 0+0= 0.
lim
h→
π
2
1−cos3h
h
49. Calculate lim
x→0−
sinx
|x|
.
solution
lim
x→0−
sinx
|x|
= lim
x→0−
sinx
−x
= −1
Usetheidentitysin3θ = 3sinθ −4sin
3
θ toevaluatethelimit lim
θ→0
sin3θ −3sinθ
θ
3
.
51. Provethefollowingresult statedinTheorem2:
lim
θ→0
1−cosθ
θ
= 0 7
Hint:
1−cosθ
θ
=
1
1+cosθ
·
1−cos
2
θ
θ
.
solution
lim
θ→0
1−cosθ
θ
= lim
θ→0
1
1+cosθ
·
1−cos
2
θ
θ
= lim
θ→0
1
1+cosθ
·
sin
2
θ
θ
= lim
θ→0
1
1+cosθ
· lim
θ→0
sin
2
θ
θ
= lim
θ→0
1
1+cosθ
· lim
θ→0
sinθ
sinθ
θ
= lim
θ→0
1
1+cosθ
· lim
θ→0
sinθ · lim
θ→0
sinθ
θ
=
1
2
· 0· 1= 0.
Investigate lim
h→0
1−cosh
h
2
numerically (andgraphically if youhaveagraphingutility). Thenprovethat the
limit isequal to
1
2
. Hint: Seethehint for Exercise51.
InExercises53–55, evaluateusingtheresult of Exercise52.
53. lim
h→0
cos3h −1
h
2
solution Wemakethesubstitutionθ = 3h. Thenh = θ/3, and
lim
h→0
cos3h −1
h
2
= lim
θ→0
cosθ −1
(θ/3)
2
= −9 lim
θ→0
1−cosθ
θ
2
= −
9
2
.
lim
h→0
cos3h −1
cos2h −1
55. lim
t →0

1−cost
t
solution lim
t →0+

1−cost
t
=
_
lim
t →0+
1−cost
t
2
=
_
1
2
=

2
2
; on the other hand, lim
t →0−

1−cost
t
=

_
lim
t →0−
1−cost
t
2
= −
_
1
2
= −

2
2
.
UsetheSqueezeTheoremtoprovethat if lim
x→c
|f (x)| = 0, then lim
x→c
f (x) = 0.
Further Insights and Challenges
57. Usetheresult of Exercise52toprovethat for m = 0,
lim
x→0
cosmx −1
x
2
= −
m
2
2
solution Substituteu = mx into
cosmx −1
x
2
. Weobtainx =
u
m
. Asx → 0, u → 0; therefore,
lim
x→0
cosmx −1
x
2
= lim
u→0
cosu −1
(u/m)
2
= lim
u→0
m
2
cosu −1
u
2
= m
2
_

1
2
_
= −
m
2
2
.
June 7, 2011 LTSV SSM Second Pass
66 C HA P T E R 2 LIMITS
Usingadiagramof theunit circleandthePythagoreanTheorem, showthat
sin
2
θ ≤ (1−cosθ)
2
+sin
2
θ ≤ θ
2
Concludethat sin
2
θ ≤ 2(1− cosθ) ≤ θ
2
andusethistogiveanalternativeproof of Eq. (7) inExercise51. Then
giveanalternativeproof of theresult inExercise52.
59. (a) Investigate lim
x→c
sinx −sinc
x −c
numericallyfor theﬁvevaluesc = 0,
π
6
,
π
4
,
π
3
,
π
2
.
(b) Canyouguesstheanswer for general c?
(c) Checkthat your answer to(b) worksfor twoother valuesof c.
solution
(a)
x c −0.01 c −0.001 c +0.001 c +0.01
sinx −sinc
x −c
0.999983 0.99999983 0.99999983 0.999983
Herec = 0andcosc = 1.
x c −0.01 c −0.001 c +0.001 c +0.01
sinx −sinc
x −c
0.868511 0.866275 0.865775 0.863511
Herec =
π
6
andcosc =

3
2
≈ 0.866025.
x c −0.01 c −0.001 c +0.001 c +0.01
sinx −sinc
x −c
0.504322 0.500433 0.499567 0.495662
Herec =
π
3
andcosc =
1
2
.
x c −0.01 c −0.001 c +0.001 c +0.01
sinx −sinc
x −c
0.710631 0.707460 0.706753 0.703559
Herec =
π
4
andcosc =

2
2
≈ 0.707107.
x c −0.01 c −0.001 c +0.001 c +0.01
sinx −sinc
x −c
0.005000 0.000500 −0.000500 −0.005000
Herec =
π
2
andcosc = 0.
(b) lim
x→c
sinx −sinc
x −c
= cosc.
(c)
x c −0.01 c −0.001 c +0.001 c +0.01
sinx −sinc
x −c
−0.411593 −0.415692 −0.416601 −0.420686
Herec = 2andcosc = cos2≈ −0.416147.
x c −0.01 c −0.001 c +0.001 c +0.01
sinx −sinc
x −c
0.863511 0.865775 0.866275 0.868511
Herec = −
π
6
andcosc =

3
2
≈ 0.866025.
2.7 Limits at Infinity
Preliminary Questions
1. Assumethat
lim
x→∞
f (x) = L and lim
x→L
g(x) = ∞
Whichof thefollowingstatementsarecorrect?
June 7, 2011 LTSV SSM Second Pass
S E C T I ON 2.7 Limits at Infinity 67
(a) x = L isavertical asymptoteof g(x).
(b) y = L isahorizontal asymptoteof g(x).
(c) x = L isavertical asymptoteof f (x).
(d) y = L isahorizontal asymptoteof f (x).
solution
(a) Because lim
x→L
g(x) = ∞, x = L isavertical asymptoteof g(x). Thisstatement iscorrect.
(b) Thisstatement isnot correct.
(c) Thisstatement isnot correct.
(d) Because lim
x→∞
f (x) = L, y = L isahorizontal asymptoteof f (x). Thisstatement iscorrect.
2. What arethefollowinglimits?
(a) lim
x→∞
x
3
(b) lim
x→−∞
x
3
(c) lim
x→−∞
x
4
solution
(a) lim
x→∞
x
3
= ∞
(b) lim
x→−∞
x
3
= −∞
(c) lim
x→−∞
x
4
= ∞
3. Sketchthegraphof afunctionthatapproachesalimitasx → ∞butdoesnotapproachalimit(eitherﬁniteorinﬁnite)
asx → −∞.
solution
y
x
4. What isthesignof a if f (x) = ax
3
+x +1satisﬁes
lim
x→−∞
f (x) = ∞?
solution Because lim
x→−∞
x
3
= −∞, a must benegativetohave lim
x→−∞
f (x) = ∞.
5. What isthesignof theleadingcoefﬁcient a
7
if f (x) isapolynomial of degree7suchthat lim
x→−∞
f (x) = ∞?
solution The behavior of f (x) as x → −∞ is controlled by the leading term; that is, lim
x→−∞
f (x) =
lim
x→−∞
a
7
x
7
. Becausex
7
→ −∞asx → −∞, a
7
must benegativetohavelim
x→−∞
f (x) = ∞.
6. Explainwhy lim
x→∞
sin
1
x
existsbut lim
x→0
sin
1
x
doesnot exist. What is lim
x→∞
sin
1
x
?
solution Asx → ∞,
1
x
→ 0, so
lim
x→∞
sin
1
x
= sin0= 0.
Ontheother hand,
1
x
→ ±∞asx → 0, andas
1
x
→ ±∞, sin
1
x
oscillatesinﬁnitelyoften. Thus
lim
x→0
sin
1
x
doesnot exist.
Exercises
1. What arethehorizontal asymptotesof thefunctioninFigure6?
−20 20 40 60 80
x
1
2
y
y = f(x)
FIGURE 6
June 7, 2011 LTSV SSM Second Pass
68 C HA P T E R 2 LIMITS
solution Because
lim
x→−∞
f (x) = 1 and lim
x→∞
f (x) = 2,
thefunctionf (x) hashorizontal asymptotesof y = 1andy = 2.
Sketchthegraphof afunctionf (x) that hasbothy = −1andy = 5ashorizontal asymptotes.
3. Sketchthegraphof afunctionf (x) withasinglehorizontal asymptotey = 3.
solution
−13
−9
−5
−1
−4 −2 2
3
y
x
Sketchthegraphsof twofunctionsf (x) andg(x) thathavebothy = −2andy = 4ashorizontal asymptotesbut
lim
x→∞
f (x) = lim
x→∞
g(x).
5. Investigatetheasymptoticbehavior of f (x) =
x
3
x
3
+x
numericallyandgraphically:
(a) Makeatableof valuesof f (x) for x = ±50, ±100, ±500, ±1000.
(b) Plot thegraphof f (x).
(c) What arethehorizontal asymptotesof f (x)?
solution
(a) Fromthetablebelow, it appearsthat
lim
x→±∞
x
3
x
3
+x
= 1.
x ±50 ±100 ±500 ±1000
f (x) 0.999600 0.999900 0.999996 0.999999
(b) Fromthegraphbelow, it alsoappearsthat
lim
x→±∞
x
3
x
3
+x
= 1.
−5 5
0.2
0.4
0.6
0.8
1.0
y
x
(c) Thehorizontal asymptoteof f (x) isy = 1.
Investigate lim
x→±∞
12x +1
_
4x
2
+9
numericallyandgraphically:
(a) Makeatableof valuesof f (x) =
12x +1
_
4x
2
+9
for x = ±100, ±500, ±1000, ±10,000.
(b) Plot thegraphof f (x).
(c) What arethehorizontal asymptotesof f (x)?
InExercises7–16, evaluatethelimit.
7. lim
x→∞
x
x +9
solution
lim
x→∞
x
x +9
= lim
x→∞
x
−1
(x)
x
−1
(x +9)
= lim
x→∞
1
1+
9
x
=
1
1+0
= 1.
lim
x→∞
3x
2
+20x
4x
2
+9
9. lim
x→∞
3x
2
+20x
2x
4
+3x
3
−29
solution
lim
x→∞
3x
2
+20x
2x
4
+3x
3
−29
= lim
x→∞
x
−4
(3x
2
+20x)
x
−4
(2x
4
+3x
3
−29)
= lim
x→∞
3
x
2
+
20
x
3
2+
3
x

29
x
4
=
0
2
= 0.
June 7, 2011 LTSV SSM Second Pass
S E C T I ON 2.7 Limits at Infinity 69
lim
x→∞
4
x +5
11. lim
x→∞
7x −9
4x +3
solution
lim
x→∞
7x −9
4x +3
= lim
x→∞
x
−1
(7x −9)
x
−1
(4x +3)
= lim
x→∞
7−
9
x
4+
3
x
=
7
4
.
lim
x→∞
9x
2
−2
6−29x
13. lim
x→−∞
7x
2
−9
4x +3
solution
lim
x→−∞
7x
2
−9
4x +3
= lim
x→−∞
x
−1
(7x
2
−9)
x
−1
(4x +3)
= lim
x→−∞
7x −
9
x
4+
3
x
= −∞.
lim
x→−∞
5x −9
4x
3
+2x +7
15. lim
x→−∞
3x
3
−10
x +4
solution
lim
x→−∞
3x
3
−10
x +4
= lim
x→−∞
x
−1
(3x
3
−10)
x
−1
(x +4)
= lim
x→−∞
3x
2

10
x
1+
4
x
=

1
= ∞.
lim
x→−∞
2x
5
+3x
4
−31x
8x
4
−31x
2
+12
InExercises17–22, ﬁndthehorizontal asymptotes.
17. f (x) =
2x
2
−3x
8x
2
+8
solution First calculatethelimitsasx → ±∞. For x → ∞,
lim
x→∞
2x
2
−3x
8x
2
+8
= lim
x→∞
2−
3
x
8+
8
x
2
=
2
8
=
1
4
.
Similarly,
lim
x→−∞
2x
2
−3x
8x
2
+8
= lim
x→−∞
2−
3
x
8+
8
x
2
=
2
8
=
1
4
.
Thus, thehorizontal asymptoteof f (x) isy =
1
4
.
f (x) =
8x
3
−x
2
7+11x −4x
4
19. f (x) =
_
36x
2
+7
9x +4
solution For x > 0, x
−1
= |x
−1
| =

x
−2
, so
lim
x→∞
_
36x
2
+7
9x +4
= lim
x→∞
_
36+
7
x
2
9+
4
x
=

36
9
=
2
3
.
Ontheother hand, for x < 0, x
−1
= −|x
−1
| = −

x
−2
, so
lim
x→−∞
_
36x
2
+7
9x +4
= lim
x→−∞

_
36+
7
x
2
9+
4
x
=

36
9
= −
2
3
.
Thus, thehorizontal asymptotesof f (x) arey =
2
3
andy = −
2
3
.
f (x) =
_
36x
4
+7
9x
2
+4
21. f (t ) =
3
t
1+3
−t
solution With
lim
t →∞
3
t
1+3
−t
=

1
= ∞
and
lim
t →−∞
3
t
1+3
−t
= 0,
thefunctionf (t ) hasonehorizontal asymptote, y = 0.
June 7, 2011 LTSV SSM Second Pass
70 C HA P T E R 2 LIMITS
f (t ) =
t
1/3
(64t
2
+9)
1/6
InExercises23–30, evaluatethelimit.
23. lim
x→∞
_
9x
4
+3x +2
4x
3
+1
solution For x > 0, x
−3
= |x
−3
| =

x
−6
, so
lim
x→∞
_
9x
4
+3x +2
4x
3
+1
= lim
x→∞
_
9
x
2
+
3
x
5
+
2
x
6
4+
1
x
3
= 0.
lim
x→∞
_
x
3
+20x
10x −2
25. lim
x→−∞
8x
2
+7x
1/3
_
16x
4
+6
solution For x < 0, x
−2
= |x
−2
| =

x
−4
, so
lim
x→−∞
8x
2
+7x
1/3
_
16x
4
+6
= lim
x→−∞
8+
7
x
5/3
_
16+
6
x
4
=
8

16
= 2.
lim
x→−∞
4x −3
_
25x
2
+4x
27. lim
t →∞
t
4/3
+t
1/3
(4t
2/3
+1)
2
solution lim
t →∞
t
4/3
+t
1/3
(4t
2/3
+1)
2
= lim
t →∞
1+
1
t
(4+
1
t
2/3
)
2
=
1
16
.
lim
t →∞
t
4/3
−9t
1/3
(8t
4
+2)
1/3
29. lim
x→−∞
|x| +x
x +1
solution For x < 0, |x| = −x. Therefore, for all x < 0,
|x| +x
x +1
=
−x +x
x +1
= 0;
consequently,
lim
x→−∞
|x| +x
x +1
= 0.
lim
t →−∞
4+6· 10
2t
5−9· 10
3t
31. Determinethelimitsat inﬁnityof g(t ) = 5
−1/t
2
.
solution Because lim
t →±∞

1
t
2
= −
1

= 0, it followsthat
lim
t →±∞
5
−1/t
2
= 5
0
= 1
Showthat lim
x→∞
(
_
x
2
+1−x) = 0. Hint: Observethat
_
x
2
+1−x =
1
_
x
2
+1+x
33. AccordingtotheMichaelis–Mentenequation(Figure7), whenanenzymeiscombinedwithasubstrateof concen-
trations (inmillimolars), thereactionrate(inmicromolars/min) is
R(s) =
As
K +s
(A, K constants)
(a) Show, bycomputing lim
s→∞
R(s), that A isthelimitingreactionrateastheconcentrations approaches∞.
(b) Showthat thereactionrateR(s) attainsone-half of thelimitingvalueA whens = K.
(c) For acertainreaction, K = 1.25mM andA = 0.1. For whichconcentrations isR(s) equal to75%of itslimiting
value?
Leonor Michaelis
1875−1949
Maud Menten
1879−1960
FIGURE 7 Canadian-bornbiochemist MaudMentenisbest knownfor her fundamental workonenzymekineticswith
Germanscientist Leonor Michaelis. Shewasalsoanaccomplishedpainter, clarinetist, mountainclimber, andmaster of
numerouslanguages.
June 7, 2011 LTSV SSM Second Pass
S E C T I ON 2.7 Limits at Infinity 71
solution
(a) lim
s→∞
R(s) = lim
s→∞
As
K +s
= lim
s→∞
A
1+
K
s
= A.
(b) Observethat
R(K) =
AK
K +K
=
AK
2K
=
A
2
,
haveof thelimitingvalue.
(c) Bypart (a), thelimitingvalueis0.1, soweneedtodeterminethevalueof s that satisﬁes
R(s) =
0.1s
1.25+s
= 0.075.
Solvingthisequationfor s yields
s =
(1.25)(0.075)
0.025
= 3.75mM.
Supposethattheaveragetemperatureof theearthisT (t ) =283+3(1−10
−0.013t
) kelvins, wheret isthenumber
of yearssince2000.
(a) Calculatethelong-termaverageL = lim
t →∞
T (t ).
(b) At what timeisT (t ) withinone-half adegreeof itslimitingvalue?
InExercises35–42, calculatethelimit.
35. lim
x→∞
_
_
4x
4
+9x −2x
2
_
solution Write
_
4x
4
+9x −2x
2
=
__
4x
4
+9x −2x
2
_
_
4x
4
+9x +2x
2
_
4x
4
+9x +2x
2
=
(4x
4
+9x) −4x
4
_
4x
4
+9x +2x
2
=
9x
_
4x
4
+9x +2x
2
.
Thus,
lim
x→∞
(
_
4x
4
+9x −2x
2
) = lim
x→∞
9x
_
4x
4
+9x +2x
2
= 0.
lim
x→∞
(
_
9x
3
+x −x
3/2
)
37. lim
x→∞
_
2

x −

x +2
_
solution Write
2

x −

x +2=
_
2

x −

x +2
_ 2

x +

x +2
2

x +

x +2
=
4x −(x +2)
2

x +

x +2
=
3x −2
2

x +

x +2
.
Thus,
lim
x→∞
(2

x −

x +2) = lim
x→∞
3x −2
2

x +

x +2
= ∞.
lim
x→∞
_
1
x

1
x +2
_
39. lim
t →∞
tan
_
π3
t
+1
4−3
t +1
_
solution Dividingnumerator anddenominator by3
t
gives
lim
t →∞
tan
_
π3
t
+1
4−3
t +1
_
= lim
t →∞
tan
_
π +3
−t
4· 3
−t
−3
_
= tan
_
π
−3
_
= tan
_

π
3
_
= −

3
lim
t →−∞
2
_
8t
t +1
−10
t +1
_
41. Let P(n) betheperimeter of ann-goninscribedinaunit circle(Figure8).
(a) Explain, intuitively, whyP(n) approaches2π asn → ∞.
(b) Showthat P(n) = 2n sin
_
π
n
_
.
(c) Combine(a) and(b) toconcludethat lim
n→∞
n
π
sin
_
π
n
_
= 1.
(d) Usethistogiveanother argument that lim
θ→0
sinθ
θ
= 1.
June 7, 2011 LTSV SSM Second Pass
72 C HA P T E R 2 LIMITS
n = 6 n = 9 n = 12
FIGURE 8
solution
(a) Asn → ∞, then-gonapproachesacircleof radius1. Therefore, theperimeter of then-gonapproachesthecircum-
ferenceof theunit circleasn → ∞. That is, P(n) → 2π asn → ∞.
(b) Eachsideof then-gonisthethirdsideof anisoscelestrianglewithequal lengthsidesof length1andangleθ =

n
betweentheequal lengthsides. Thelengthof eachsideof then-gonistherefore
_
1
2
+1
2
−2cos

n
=
_
2(1−cos

n
) =
_
4sin
2
π
n
= 2sin
π
n
.
Finally,
P(n) = 2n sin
π
n
.
(c) Combiningparts(a) and(b),
lim
n→∞
P(n) = lim
n→∞
2n sin
π
n
= 2π.
Dividingbothsidesof thislast expressionby2π yields
lim
n→∞
n
π
sin
π
n
= 1.
(d) Let θ =
π
n
. Thenθ → 0asn → ∞,
n
π
sin
π
n
=
1
θ
sinθ =
sinθ
θ
,
and
lim
n→∞
n
π
sin
π
n
= lim
θ→0
sinθ
θ
= 1.
PhysicistshaveobservedthatEinstein’stheoryof special relativityreducestoNewtonianmechanicsinthelimit
as c → ∞, wherec is thespeed of light. This is illustrated by astonetossed up vertically fromground level so
that it returnstoearthonesecondlater. UsingNewton’sLaws, weﬁndthat thestone’smaximumheight ish = g/8
meters(g = 9.8m/s
2
). Accordingtospecial relativity, thestone’smassdependsonitsvelocitydividedbyc, andthe
maximumheight is
h(c) = c
_
c
2
/g
2
+1/4−c
2
/g
Provethat lim
c→∞
h(c) = g/8.
Further Insights and Challenges
43. Everylimit asx → ∞canberewrittenasaone-sidedlimit ast → 0+, wheret = x
−1
. Settingg(t ) = f (t
−1
), we
have
lim
x→∞
f (x) = lim
t →0+
g(t )
Showthat lim
x→∞
3x
2
−x
2x
2
+5
= lim
t →0+
3−t
2+5t
2
, andevaluateusingtheQuotient Law.
solution Let t = x
−1
. Thenx = t
−1
, t → 0+ asx → ∞, and
3x
2
−x
2x
2
+5
=
3t
−2
−t
−1
2t
−2
+5
=
3−t
2+5t
2
.
Thus,
lim
x→∞
3x
2
−x
2x
2
+5
= lim
t →0+
3−t
2+5t
2
=
3
2
.
Rewritethefollowingasone-sidedlimitsasinExercise43andevaluate.
(a) lim
x→∞
3−12x
3
4x
3
+3x +1
(b) lim
x→∞
2
1/x
(c) lim
x→∞
x sin
1
x
(d) lim
x→∞
cos
_
π
x +1
x −1
_
45. Let G(b) = lim
x→∞
(1+b
x
)
1/x
for b ≥ 0. InvestigateG(b) numerically and graphically for b = 0.2, 0.8, 2, 3, 5
(and additional values if necessary). Then makeaconjecturefor thevalueof G(b) as afunction of b. Drawagraph
of y = G(b). Does G(b) appear to becontinuous? Wewill evaluateG(b) using L’Hôpital’s Rulein Section 7.7(see
Exercise65inSection7.7).
June 7, 2011 LTSV SSM Second Pass
S E C T I ON 2.8 Intermediate Value Theorem 73
solution
• b = 0.2:
x 5 10 50 100
f (x) 1.000064 1.000000 1.000000 1.000000
It appearsthat G(0.2) = 1.
• b = 0.8:
x 5 10 50 100
f (x) 1.058324 1.010251 1.000000 1.000000
It appearsthat G(0.8) = 1.
• b = 2:
x 5 10 50 100
f (x) 2.012347 2.000195 2.000000 2.000000
It appearsthat G(2) = 2.
• b = 3:
x 5 10 50 100
f (x) 3.002465 3.000005 3.000000 3.000000
It appearsthat G(3) = 3.
• b = 5:
x 5 10 50 100
f (x) 5.000320 5.000000 5.000000 5.000000
It appearsthat G(5) = 5.
Basedontheseobservationsweconjecturethat G(b) = 1if 0≤ b ≤ 1andG(b) = b for b > 1. Thegraphof y = G(b)
isshownbelow; thegraphdoesappear tobecontinuous.
1
0
0 1 2 3 4
2
3
4
y
x
2.8 Intermediate Value Theorem
Preliminary Questions
1. Provethat f (x) = x
2
takesonthevalue0.5intheinterval [0, 1].
solution Observethat f (x) = x
2
iscontinuouson[0, 1] withf (0) = 0andf (1) = 1. Becausef (0) < 0.5< f (1),
theIntermediateValueTheoremguaranteesthereisac ∈ [0, 1] suchthat f (c) = 0.5.
2. ThetemperatureinVancouver was8

Cat6am androseto20

Catnoon. Whichassumptionabouttemperatureallows
ustoconcludethat thetemperaturewas15

C at somemoment of timebetween6am andnoon?
solution Wemust assumethat temperatureisacontinuousfunctionof time.
3. What isthegraphical interpretationof theIVT?
solution If f iscontinuouson[a, b], thenthehorizontal liney = k for everykbetweenf (a) andf (b) intersectsthe
graphof y = f (x) at least once.
June 7, 2011 LTSV SSM Second Pass
74 C HA P T E R 2 LIMITS
4. Showthat thefollowingstatement isfalsebydrawingagraphthat providesacounterexample:
If f (x) iscontinuousandhasaroot in[a, b], thenf (a) andf (b) haveoppositesigns.
solution
f(a)
f(b)
a
y
x
b
5. Assumethat f (t ) iscontinuouson[1, 5] andthat f (1) = 20, f (5) = 100. Determinewhether eachof thefollowing
statementsisalwaystrue, never true, or sometimestrue.
(a) f (c) = 3hasasolutionwithc ∈ [1, 5].
(b) f (c) = 75hasasolutionwithc ∈ [1, 5].
(c) f (c) = 50hasnosolutionwithc ∈ [1, 5].
(d) f (c) = 30hasexactlyonesolutionwithc ∈ [1, 5].
solution
(a) Thisstatement issometimestrue.
(b) Thisstatement isalwaystrue.
(c) Thisstatement isnever true.
(d) Thisstatement issometimestrue.
Exercises
1. UsetheIVT toshowthat f (x) = x
3
+x takesonthevalue9for somex in[1, 2].
solution Observethat f (1) = 2andf (2) = 10. Sincef isapolynomial, it iscontinuouseverywhere; inparticular
on[1, 2]. Therefore, bytheIVT thereisac ∈ [1, 2] suchthat f (c) = 9.
Showthat g(t ) =
t
t +1
takesonthevalue0.499for somet in[0, 1].
3. Showthat g(t ) = t
2
tant takesonthevalue
1
2
for somet in
_
0,
π
4
_
.
solution g(0) = 0andg(
π
4
) =
π
2
16
. g(t ) iscontinuousfor all t between0and
π
4
, and0<
1
2
<
π
2
16
; therefore, bythe
IVT, thereisac ∈ [0,
π
4
] suchthat g(c) =
1
2
.
Showthat f (x) =
x
2
x
7
+1
takesonthevalue0.4.
5. Showthat cosx = x hasasolutionintheinterval [0, 1]. Hint: Showthat f (x) = x −cosx hasazeroin[0, 1].
solution Let f (x) = x − cosx. Observethat f is continuous with f (0) = −1 and f (1) = 1− cos1 ≈ 0.46.
Therefore, by theIVT thereis ac ∈ [0, 1] such that f (c) = c − cosc = 0. Thus c = cosc and hencetheequation
cosx = x hasasolutionc in[0, 1].
UsetheIVT toﬁndaninterval of length
1
2
containingaroot of f (x) = x
3
+2x +1.
InExercises7–16, proveusingtheIVT.
7.

c +

c +2= 3hasasolution.
solution Let f (x) =

x +

x +2− 3. Notethat f is continuous on
_
1
4
, 2
_
withf (
1
4
) =
_
1
4
+
_
9
4
− 3 = −1
andf (2) =

2−1≈ 0.41. Therefore, bytheIVT thereisac ∈
_
1
4
, 2
_
suchthat f (c) =

c +

c +2−3= 0. Thus

c +

c +2= 3andhencetheequation

x +

x +2= 3hasasolutionc in
_
1
4
, 2
_
.
For all integersn, sinnx = cosx for somex ∈ [0, π].
9.

2exists. Hint: Consider f (x) = x
2
.
solution Letf (x) = x
2
. Observethatf iscontinuouswithf (1) = 1andf (2) = 4. Therefore, bytheIVT thereisa
c ∈ [1, 2] suchthat f (c) = c
2
= 2. Thisprovestheexistenceof

2, anumber whosesquareis2.
A positivenumber c hasannthroot for all positiveintegersn.
11. For all positiveintegersk, cosx = x
k
hasasolution.
solution For eachpositiveinteger k, letf (x) = x
k
−cosx. Observethatf iscontinuouson
_
0,
π
2
_
withf (0) = −1
andf (
π
2
) =
_
π
2
_
k
> 0. Therefore, bytheIVT thereisac ∈
_
0,
π
2
_
suchthat f (c) = c
k
−cos(c) = 0. Thuscosc = c
k
andhencetheequationcosx = x
k
hasasolutionc intheinterval
_
0,
π
2
_
.
2
x
= bx hasasolutionif b > 2.
13. 2
x
+3
x
= 4
x
hasasolution.
solution Let f (x) = 2
x
+3
x
−4
x
. Observethat f iscontinuouson[0, 2] withf (0) = 1> 0andf (2) = −3< 0.
Therefore, bytheIVT, thereisac ∈ (0, 2) suchthat f (c) = 2
c
+3
c
−4
c
= 0.
tanx = x hasinﬁnitelymanysolutions.
June 7, 2011 LTSV SSM Second Pass
S E C T I ON 2.8 Intermediate Value Theorem 75
15. 2
x
+
1
x
= −4hasasolution.
solution Let f (x) = 2
x
+
1
x
+ 4. Observethat f iscontinuousfor x < 0withf (−1) = 2
−1
+
1
−1
+ 4=
7
2
> 0
and f
_

1
8
_
= 2
−1/8
− 8+ 4 ≈ −3.08 < 0. Therefore, by the IVT, there is a c ∈
_
−1, −
1
8
_
such that f (c) =
2
c

1
c
+4= 0andthus2
c

1
c
= −4.
x = sinx +cosx hasasolution.
17. Carryout threestepsof theBisectionMethodfor f (x) = 2
x
−x
3
asfollows:
(a) Showthat f (x) hasazeroin[1, 1.5].
(b) Showthat f (x) hasazeroin[1.25, 1.5].
(c) Determinewhether [1.25, 1.375] or [1.375, 1.5] containsazero.
solution Notethat f (x) iscontinuousfor all x.
(a) f (1) = 1, f (1.5) = 2
1.5
−(1.5)
3
< 3−3.375< 0. Hence, f (x) = 0for somex between1and1.5.
(b) f (1.25) ≈ 0.4253> 0andf (1.5) < 0. Hence, f (x) = 0for somex between1.25and1.5.
(c) f (1.375) ≈ −0.0059. Hence, f (x) = 0for somex between1.25and1.375.
Figure4showsthat f (x) = x
3
−8x −1hasaroot intheinterval [2.75, 3]. ApplytheBisectionMethodtwice
toﬁndaninterval of length
1
16
containingthisroot.
19. Findaninterval of length
1
4
in[1, 2] containingaroot of theequationx
7
+3x −10= 0.
solution Let f (x) = x
7
+ 3x − 10. Observethat f is continuous with f (1) = −6 and f (2) = 124. Therefore,
by the IVT there is a c ∈ [1, 2] such that f (c) = 0. f (1.5) ≈ 11.59 > 0, so f (c) = 0 for some c ∈ [1, 1.5].
f (1.25) ≈ −1.48< 0, andsof (c) = 0for somec ∈ [1.25, 1.5]. Thismeansthat[1.25, 1.5] isaninterval of length0.25
containingaroot of f (x).
Showthattan
3
θ −8tan
2
θ +17tanθ −8= 0hasarootin[0.5, 0.6]. ApplytheBisectionMethodtwicetoﬁnd
aninterval of length0.025containingthisroot.
InExercises21–24, drawthegraphof afunctionf (x) on[0, 4] withthegivenproperty.
21. J umpdiscontinuityat x = 2anddoesnot satisfytheconclusionof theIVT.
solution Thefunctiongraphedbelowhas ajumpdiscontinuity at x = 2. Notethat whilef (0) = 2andf (4) = 4,
thereisnopoint c intheinterval [0, 4] suchthat f (c) = 3. Accordingly, theconclusionof theIVT isnot satisﬁed.
4 3 2 1
1
3
2
4
y
x
J umpdiscontinuityat x = 2andsatisﬁestheconclusionof theIVT on[0, 4].
23. Inﬁniteone-sidedlimitsat x = 2anddoesnot satisfytheconclusionof theIVT.
solution Thefunctiongraphedbelowhasinﬁniteone-sidedlimitsat x = 2. Notethat whilef (0) = 2andf (4) = 4,
thereisnopoint c intheinterval [0, 4] suchthat f (c) = 3. Accordingly, theconclusionof theIVT isnot satisﬁed.
4 3 2 1
1
−1
3
2
4
5
6
y
x
Inﬁniteone-sidedlimitsat x = 2andsatisﬁestheconclusionof theIVT on[0, 4].
25. CanCorollary2beappliedtof (x) = x
−1
on[−1, 1]? Doesf (x) haveanyroots?
solution No, becausef (x) = x
−1
isnot continuouson[−1, 1]. Eventhoughf (−1) = −1< 0andf (1) = 1> 0,
thefunctionhasnorootsbetweenx = −1andx = 1. Infact, thisfunctionhasnorootsat all.
June 7, 2011 LTSV SSM Second Pass
76 C HA P T E R 2 LIMITS
Further Insights and Challenges
Takeany mapanddrawacircleonit anywhere(Figure5). Provethat at any moment intimethereexistsapair
of diametrically oppositepoints A andB onthat circlecorrespondingto locations wherethetemperatures at that
momentareequal. Hint: Letθ beanangularcoordinatealongthecircleandletf (θ) bethedifferenceintemperatures
at thelocationscorrespondingtoθ andθ +π.
27. Showthat if f (x) is continuous and 0 ≤ f (x) ≤ 1 for 0 ≤ x ≤ 1, then f (c) = c for somec in [0, 1]
(Figure6).
1
1
y = f(x)
y = x
c
x
y
FIGURE 6 A functionsatisfying0≤ f (x) ≤ 1for 0≤ x ≤ 1.
solution If f (0) = 0, the proof is done with c = 0. We may assume that f (0) > 0. Let g(x) = f (x) − x.
g(0) = f (0) − 0 = f (0) > 0. Sincef (x) is continuous, theRuleof Differences dictates that g(x) is continuous. We
needtoprovethat g(c) = 0for somec ∈ [0, 1]. Sincef (1) ≤ 1, g(1) = f (1) − 1≤ 0. If g(1) = 0, theproof isdone
withc = 1, solet’sassumethat g(1) < 0.
Wenowhaveacontinuousfunctiong(x) ontheinterval [0, 1] suchthat g(0) > 0andg(1) < 0. FromtheIVT, there
must besomec ∈ [0, 1] sothat g(c) = 0, sof (c) −c = 0andsof (c) = c.
Thisisasimplecaseof averygeneral, useful, andbeautiful theoremcalledtheBrouwer ﬁxedpointtheorem.
UsetheIVTtoshowthatif f (x) iscontinuousandone-to-oneonaninterval [a, b], thenf (x) iseitheranincreasing
or adecreasingfunction.
29. HamSandwichTheoremFigure7(A) shows asliceof ham. Provethat for any angleθ (0 ≤ θ ≤ π), it
is possibleto cut theslicein half with acut of inclineθ. Hint: Thelines of inclination θ aregiven by theequations
y = (tanθ)x +b, whereb variesfrom−∞to∞. Eachsuchlinedividesthesliceintotwopieces(oneof whichmaybe
empty). Let A(b) betheamount of hamtotheleft of thelineminustheamount totheright, andlet A bethetotal areaof
theham. Showthat A(b) = −A if b is sufﬁciently largeandA(b) = A if b is sufﬁciently negative. ThenusetheIVT.
Thisworksif θ = 0or
π
2
. If θ = 0, deﬁneA(b) astheamount of hamabovetheliney = b minustheamount below.
Howcanyoumodifytheargument toworkwhenθ =
π
2
(inwhichcasetanθ = ∞)?
Cutting asliceof ham
at an angle .
L(0) = L( )
L( )
L( )
2
(A) (B) A sliceof hamon top
of asliceof bread.
x
y
x
y
FIGURE 7
solution Let θ besuchthat θ =
π
2
. For anyb, consider thelineL(θ) drawnat angleθ tothex axisstartingat (0, b).
Thislinehasformulay = (tanθ)x +b. Let A(b) betheamount of hamabovethelineminusthat belowtheline.
Let A > 0betheareaof theham. Wehavetoaccept thefollowing(reasonable) assumptions:
• For lowenoughb = b
0
, thelineL(θ) liesentirelybelowtheham, sothat A(b
0
) = A −0= A.
• For highenoughb
1
, thelineL(θ) liesentirelyabovetheham, sothat A(b
1
) = 0−A = −A.
• A(b) iscontinuousasafunctionof b.
Under theseassumptions, weseeA(b) isacontinuousfunctionsatisfyingA(b
0
) > 0andA(b
1
) < 0for someb
0
< b
1
.
BytheIVT, A(b) = 0for someb ∈ [b
0
, b
1
].
Supposethat θ =
π
2
. Let thelineL(c) bethevertical linethrough(c, 0) (x = c). Let A(c) betheareaof hamtothe
left of L(c) minus that to theright of L(c). SinceL(0) lies entirely to theleft of theham, A(0) = 0− A = −A. For
somec = c
1
sufﬁciently large, L(c) liesentirely totheright of theham, sothat A(c
1
) = A − 0 = A. HenceA(c) isa
continuousfunctionof c suchthat A(0) < 0andA(c
1
) > 0. BytheIVT, thereissomec ∈ [0, c
1
] suchthat A(c) = 0.
Figure7(B) showsasliceof hamonapieceof bread. Provethatitispossibletoslicethisopen-facedsandwich
sothat eachpart hasequal amountsof hamandbread. Hint: ByExercise29, for all 0≤ θ ≤ π thereisalineL(θ)
of inclineθ (whichweassumeisunique) that dividesthehamintotwoequal pieces. Let B(θ) denotetheamount of
breadtotheleftof (or above) L(θ) minustheamounttotheright(or below). NoticethatL(π) andL(0) arethesame
line, but B(π) = −B(0) sinceleft andright get interchangedas theanglemoves from0to π. Assumethat B(θ)
is continuous andapply theIVT. (By afurther extensionof this argument, onecanprovethefull “HamSandwich
Theorem,” whichstatesthat if youallowtheknifetocut at aslant, thenit ispossibletocut asandwichconsistingof
asliceof hamandtwoslicesof breadsothat all threelayersaredividedinhalf.)
2.9 The Formal Definition of a Limit
Preliminary Questions
1. Giventhat lim
x→0
cosx = 1, whichof thefollowingstatementsistrue?
(a) If |cosx −1| isverysmall, thenx iscloseto0.
(b) Thereisan > 0suchthat |x| < 10
−5
if 0< |cosx −1| < .
June 7, 2011 LTSV SSM Second Pass
S E C T I ON 2.9 The Formal Definition of a Limit 77
(c) Thereisaδ > 0suchthat |cosx −1| < 10
−5
if 0< |x| < δ.
(d) Thereisaδ > 0suchthat |cosx| < 10
−5
if 0< |x −1| < δ.
solution Thetruestatement is(c): Thereisaδ > 0suchthat |cosx −1| < 10
−5
if 0< |x| < δ.
2. Supposeit isknownthat for agiven andδ, |f (x) − 2| < if 0< |x − 3| < δ. Whichof thefollowingstatements
must alsobetrue?
(a) |f (x) −2| < if 0< |x −3| < 2δ
(b) |f (x) −2| < 2 if 0< |x −3| < δ
(c) |f (x) −2| <

2
if 0< |x −3| <
δ
2
(d) |f (x) −2| < if 0< |x −3| <
δ
2
solution Statements(b)and(d)aretrue.
Exercises
1. BasedontheinformationconveyedinFigure5(A), ﬁndvaluesof L, , andδ > 0suchthat thefollowingstatement
holds: |f (x) −L| < if 0< |x| < δ.
3 3.1 2.9
10
10.4
9.8
x
y
y = f(x)
y = f(x)
(A) (B)
0.1 −0.1
4
4.8
3.5
x
y
FIGURE 5
solution Wesee−0.1 < x < 0.1forces 3.5 < f (x) < 4.8. Rewritten, this means that |x − 0| < 0.1implies that
|f (x) −4| < 0.8. Replacingnumberswhereappropriateinthedeﬁnitionof thelimit|x −c| < δ implies|f (x) −L| < ,
weget L = 4, = 0.8, c = 0, andδ = 0.1.
Based on theinformation conveyed in Figure5(B), ﬁnd values of c, L, , and δ > 0 such that thefollowing
statement holds: |f (x) −L| < if 0< |x −c| < δ.
3. Consider lim
x→4
f (x), wheref (x) = 8x +3.
(a) Showthat |f (x) −35| = 8|x −4|.
(b) Showthat for any > 0, |f (x) −35| < if 0< |x −4| < δ, whereδ =

8
. Explainhowthisprovesrigorouslythat
lim
x→4
f (x) = 35.
solution
(a) |f (x) −35| = |8x +3−35| = |8x −32| = |8(x −4)| = 8|x −4|. (Rememberthatthelaststepisjustiﬁedbecause
8> 0).
(b) Let > 0. Let δ = /8andsuppose|x − 4| < δ. By part (a), |f (x) −35| = 8|x −4| < 8δ. Substitutingδ = /8,
wesee|f (x) −35| < 8/8= . Weseethat, for any > 0, wefound an appropriateδ so that |x − 4| < δ implies
|f (x) −35| < . Hence lim
x→4
f (x) = 35.
Consider lim
x→2
f (x), wheref (x) = 4x −1.
(a) Showthat |f (x) −7| < 4δ if 0< |x −2| < δ.
(b) Findaδ suchthat
|f (x) −7| < 0.01 if 0< |x −2| < δ
(c) Proverigorouslythat lim
x→2
f (x) = 7.
5. Consider lim
x→2
x
2
= 4(refer toExample2).
(a) Showthat |x
2
−4| < 0.05if 0< |x −2| < 0.01.
(b) Showthat |x
2
−4| < 0.0009if 0< |x −2| < 0.0002.
(c) Findavalueof δ suchthat |x
2
−4| islessthan10
−4
if
0< |x −2| < δ.
solution
(a) If 0< |x −2| < δ = 0.01, then|x| < 3and
¸
¸
¸x
2
−4
¸
¸
¸ = |x −2||x +2| ≤ |x −2| (|x| +2) < 5|x −2| < 0.05.
(b) If 0< |x −2| < δ = 0.0002, then|x| < 2.0002and
¸
¸
¸x
2
−4
¸
¸
¸ = |x −2||x +2| ≤ |x −2| (|x| +2) < 4.0002|x −2| < 0.00080004< 0.0009.
June 7, 2011 LTSV SSM Second Pass
78 C HA P T E R 2 LIMITS
(c) Notethat
¸
¸
¸x
2
−4
¸
¸
¸ = |(x +2)(x −2)| ≤ |x +2| |x −2|. Since|x −2| can get arbitrarily small, wecan require
|x −2| < 1so that 1 < x < 3. This ensures that |x +2| is at most 5. Nowweknowthat
¸
¸
¸x
2
−4
¸
¸
¸ ≤ 5|x − 2|. Let
δ = 10
−5
. Then, if |x −2| < δ, weget
¸
¸
¸x
2
−4
¸
¸
¸ ≤ 5|x −2| < 5×10
−5
< 10
−4
asdesired.
Withregardtothelimit lim
x→5
x
2
= 25,
(a) Showthat |x
2
−25| < 11|x −5| if 4< x < 6. Hint: Write|x
2
−25| = |x +5| · |x −5|.
(b) Findaδ suchthat |x
2
−25| < 10
−3
if 0< |x −5| < δ.
(c) Givearigorousproof of thelimitbyshowingthat|x
2
−25| < if 0< |x −5| < δ, whereδ isthesmaller of

11
and1.
7. Refer toExample3toﬁndavalueof δ > 0suchthat
¸
¸
¸
¸
1
x

1
3
¸
¸
¸
¸
< 10
−4
if 0< |x −3| < δ
solution TheExampleshowsthat for any > 0wehave
¸
¸
¸
¸
1
x

1
3
¸
¸
¸
¸
≤ if |x −3| < δ
whereδ isthesmaller of thenumbers6 and1. Inour case, wemaytakeδ = 6×10
−4
.
UseFigure6toﬁndavalueof δ > 0suchthat thefollowingstatement holds:
¸
¸
1/x
2

1
4
¸
¸
< if 0< |x −2| < δ
for = 0.03. Thenﬁndavalueof δ that worksfor = 0.01.
9. Plot f (x) =

2x −1together withthehorizontal linesy = 2.9andy = 3.1. Usethisplot toﬁndavalueof
δ > 0suchthat |

2x −1−3| < 0.1if 0< |x −5| < δ.
solution Fromtheplot below, weseethat δ = 0.25will guaranteethat |

2x −1−3| < 0.1whenever |x −5| ≤ δ.
4.6 4.8 5 5.2 5.4
2.9
2.8
3
3.1
x
y
Plot f (x) = tanx together withthehorizontal lines y = 0.99andy = 1.01. Usethis plot toﬁndavalueof
δ > 0suchthat |tanx −1| < 0.01if 0<
¸
¸
x −
π
4
¸
¸
< δ.
11. The function f (x) = 2
−x
2
satisﬁes lim
x→0
f (x) = 1. Use a plot of f to ﬁnd a value of δ > 0 such that
|f (x) −1| < 0.001if 0< |x| < δ.
solution Fromtheplot below, weseethat δ = 0.03will guaranteethat
¸
¸
¸2
−x
2
−1
¸
¸
¸ < 0.001
whenever 0< |x| < δ.
0.9985
−0.04 −0.02
0
0.02 0.04
x
y
0.9990
0.9995
1.0000
Let f (x) =
4
x
2
+1
and = 0.5. Usingaplot of f (x), ﬁndavalueof δ > 0suchthat
¸
¸
¸f (x) −
16
5
¸
¸
¸ < for
0<
¸
¸
¸x −
1
2
¸
¸
¸ < δ. Repeat for = 0.2and0.1.
13. Consider lim
x→2
1
x
.
(a) Showthat if |x −2| < 1, then
¸
¸
¸
¸
1
x

1
2
¸
¸
¸
¸
<
1
2
|x −2|
(b) Let δ bethesmaller of 1and2. Prove:
¸
¸
¸
¸
1
x

1
2
¸
¸
¸
¸
< if 0< |x −2| < δ
(c) Findaδ > 0suchthat
¸
¸
¸
1
x

1
2
¸
¸
¸ < 0.01if 0< |x −2| < δ.
(d) Proverigorouslythat lim
x→2
1
x
=
1
2
.
June 7, 2011 LTSV SSM Second Pass
S E C T I ON 2.9 The Formal Definition of a Limit 79
solution
(a) Since|x −2| < 1, it followsthat 1< x < 3, inparticular that x > 1. Becausex > 1, then
1
x
< 1and
¸
¸
¸
¸
1
x

1
2
¸
¸
¸
¸
=
¸
¸
¸
¸
2−x
2x
¸
¸
¸
¸
=
|x −2|
2x
<
1
2
|x −2|.
(b) Let δ = min{1, 2} andsupposethat |x −2| < δ. Thenbypart (a) wehave
¸
¸
¸
¸
1
x

1
2
¸
¸
¸
¸
<
1
2
|x −2| <
1
2
δ <
1
2
· 2 = .
(c) Chooseδ = 0.02. Then
¸
¸
¸
¸
1
x

1
2
¸
¸
¸
¸
<
1
2
δ = 0.01bypart (b).
(d) Let > 0begiven. Thenwhenever 0< |x −2| < δ = min{1, 2}, wehave
¸
¸
¸
¸
1
x

1
2
¸
¸
¸
¸
<
1
2
δ ≤ .
Since wasarbitrary, weconcludethat lim
x→2
1
x
=
1
2
.
Consider lim
x→1

x +3.
(a) Showthat |

x +3−2| <
1
2
|x −1| if |x −1| < 4. Hint: Multiplytheinequalityby|

x +3+2| andobserve
that |

x +3+2| > 2.
(b) Findδ > 0suchthat |

x +3−2| < 10
−4
for 0< |x −1| < δ.
(c) Proverigorouslythat thelimit isequal to2.
15. Let f (x) = sinx. Usingacalculator, weﬁnd:
f
_
π
4
−0.1
_
≈ 0.633, f
_
π
4
_
≈ 0.707, f
_
π
4
+0.1
_
≈ 0.774
Usethesevaluesandthefact that f (x) isincreasingon
_
0,
π
2
_
tojustifythestatement
¸
¸
¸f (x) −f
_
π
4

¸
¸ < 0.08 if 0<
¸
¸
¸x −
π
4
¸
¸
¸ < 0.1
ThendrawaﬁgurelikeFigure3toillustratethisstatement.
solution Sincef (x) isincreasingontheinterval, thethreef (x) valuestell usthat 0.633≤ f (x) ≤ 0.774for all x
between
π
4
−0.1and
π
4
+0.1. Wemaysubtract f (
π
4
) fromtheinequalityfor f (x). Thisshowthat, for
π
4
−0.1< x <
π
4
+ 0.1, 0.633− f (
π
4
) ≤ f (x) − f (
π
4
) ≤ 0.774− f (
π
4
). Thismeansthat, if |x −
π
4
| < 0.1, then0.633− 0.707 ≤
f (x) − f (
π
4
) ≤ 0.774− 0.707, so −0.074 ≤ f (x) − f (
π
4
) ≤ 0.067. Then−0.08 < f (x) − f (
π
4
) < 0.08follows
fromthis, so|x −
π
4
| < 0.1implies|f (x) −f (
π
4
)| < 0.08. Theﬁgurebelowillustratesthis.
0.25 0.5 0.75 1 1.25 1.5
1
0.8
0.6
0.4
0.2
x
y
Adapt theargument inExample1toproverigorouslythat lim
x→c
(ax +b) = ac +b, wherea, b, c arearbitrary.
17. Adapt theargument inExample2toproverigorouslythat lim
x→c
x
2
= c
2
for all c.
solution Torelatethegapto|x −c|, wetake
¸
¸
¸x
2
−c
2
¸
¸
¸ = |(x +c)(x −c)| = |x +c| |x −c| .
Wechooseδ intwosteps. First, sincewearerequiring|x − c| tobesmall, werequireδ < |c|, sothat x liesbetween0
and2c. Thismeansthat |x +c| < 3|c|, so|x −c||x +c| < 3|c|δ. Next, werequirethat δ <

3|c|
, so
|x −c||x +c| <

3|c|
3|c| = ,
andwearedone.
Therefore, given > 0, welet
δ = min
_
|c|,

3|c|
_
.
Then, for |x −c| < δ, wehave
|x
2
−c
2
| = |x −c| |x +c| < 3|c|δ < 3|c|

3|c|
= .
June 7, 2011 LTSV SSM Second Pass
80 C HA P T E R 2 LIMITS
Adapt theargument inExample3toproverigorouslythat lim
x→c
x
−1
=
1
c
for all c = 0.
InExercises19–24, usetheformal deﬁnitionof thelimit toprovethestatement rigorously.
19. lim
x→4

x = 2
solution Let > 0begiven. Webound|

x −2| bymultiplying

x +2

x +2
.
|

x −2| =
¸
¸
¸
¸

x −2
_√
x +2

x +2

¸
¸
¸
=
¸
¸
¸
¸
x −4

x +2
¸
¸
¸
¸
= |x −4|
¸
¸
¸
¸
1

x +2
¸
¸
¸
¸
.
Wecanassumeδ < 1, sothat |x −4| < 1, andhence

x +2>

3+2> 3. Thisgivesus
|

x −2| = |x −4|
¸
¸
¸
¸
1

x +2
¸
¸
¸
¸
< |x −4|
1
3
.
Let δ = min(1, 3). If |x −4| < δ,
|

x −2| = |x −4|
¸
¸
¸
¸
1

x +2
¸
¸
¸
¸
< |x −4|
1
3
< δ
1
3
< 3
1
3
= ,
thusprovingthelimit rigorously.
lim
x→1
(3x
2
+x) = 4
21. lim
x→1
x
3
= 1
solution Let > 0begiven. Webound
¸
¸
¸x
3
−1
¸
¸
¸ byfactoringthedifferenceof cubes:
¸
¸
¸x
3
−1
¸
¸
¸ =
¸
¸
¸(x
2
+x +1)(x −1)
¸
¸
¸ = |x −1|
¸
¸
¸x
2
+x +1
¸
¸
¸ .
Let δ = min(1,

7
), andassume|x − 1| < δ. Sinceδ < 1, 0 < x < 2. Sincex
2
+ x + 1increases as x increases for
x > 0, x
2
+x +1< 7for 0< x < 2, andso
¸
¸
¸x
3
−1
¸
¸
¸ = |x −1|
¸
¸
¸x
2
+x +1
¸
¸
¸ < 7|x −1| < 7

7
=
andthelimit isrigorouslyproven.
lim
x→0
(x
2
+x
3
) = 0
23. lim
x→2
x
−2
=
1
4
solution Let > 0begiven. First, weboundx
−2

1
4
:
¸
¸
¸
¸
x
−2

1
4
¸
¸
¸
¸
=
¸
¸
¸
¸
¸
4−x
2
4x
2
¸
¸
¸
¸
¸
= |2−x|
¸
¸
¸
¸
2+x
4x
2
¸
¸
¸
¸
.
Let δ = min(1,
4
5
), andsuppose|x − 2| < δ. Sinceδ < 1, |x − 2| < 1, so1 < x < 3. Thismeansthat 4x
2
> 4and
|2+x| < 5, sothat
2+x
4x
2
<
5
4
. Weget:
¸
¸
¸
¸
x
−2

1
4
¸
¸
¸
¸
= |2−x|
¸
¸
¸
¸
2+x
4x
2
¸
¸
¸
¸
<
5
4
|x −2| <
5
4
·
4
5
= .
andthelimit isrigorouslyproven.
lim
x→0
x sin
1
x
= 0
25. Letf (x) =
x
|x|
. Proverigorouslythat lim
x→0
f (x) doesnotexist. Hint: Showthatfor anyL, therealwaysexistssome
x suchthat |x| < δ but |f (x) −L| ≥
1
2
, nomatter howsmall δ istaken.
solution Let L beanyreal number. Let δ > 0beanysmall positivenumber. Let x =
δ
2
, whichsatisﬁes|x| < δ, and
f (x) = 1. Weconsider twocases:
• (|f (x) −L| ≥
1
2
) : wearedone.
• (|f (x) −L| <
1
2
): Thismeans
1
2
< L <
3
2
. Inthiscase, let x = −
δ
2
. f (x) = −1, andso
3
2
< L −f (x).
Ineither case, thereexistsanx suchthat |x| <
δ
2
, but |f (x) −L| ≥
1
2
.
Proverigorouslythat lim
x→0
|x| = 0.
June 7, 2011 LTSV SSM Second Pass
S E C T I ON 2.9 The Formal Definition of a Limit 81
27. Let f (x) = min(x, x
2
), wheremin(a, b) istheminimumof a andb. Proverigorouslythat lim
x→1
f (x) = 1.
solution Let > 0andlet δ = min(1,

2
). Then, whenever |x − 1| < δ, it follows that 0 < x < 2. If 1 < x < 2,
thenmin(x, x
2
) = x and
|f (x) −1| = |x −1| < δ <

2
< .
Ontheother hand, if 0< x < 1, thenmin(x, x
2
) = x
2
, |x +1| < 2and
|f (x) −1| = |x
2
−1| = |x −1| |x +1| < 2δ < .
Thus, whenever |x −1| < δ, |f (x) −1| < .
Proverigorouslythat lim
x→0
sin
1
x
doesnot exist.
29. First, usetheidentity
sinx +siny = 2sin
_
x +y
2
_
cos
_
x −y
2
_
toverifytherelation
sin(a +h) −sina = h
sin(h/2)
h/2
cos
_
a +
h
2
_
6
Thenusetheinequality
¸
¸
¸
¸
sinx
x
¸
¸
¸
¸
≤ 1for x = 0toshowthat |sin(a +h) −sina| < |h| for all a. Finally, proverigorously
that lim
x→a
sinx = sina.
solution Weﬁrst write
sin(a +h) −sina = sin(a +h) +sin(−a).
Applyingtheidentitywithx = a +h, y = −a, yields:
sin(a +h) −sina = sin(a +h) +sin(−a) = 2sin
_
a +h −a
2
_
cos
_
2a +h
2
_
= 2sin
_
h
2
_
cos
_
a +
h
2
_
= 2
_
h
h
_
sin
_
h
2
_
cos
_
a +
h
2
_
= h
sin(h/2)
h/2
cos
_
a +
h
2
_
.
Therefore,
|sin(a +h) −sina| = |h|
¸
¸
¸
¸
sin(h/2)
h/2
¸
¸
¸
¸
¸
¸
¸
¸
cos
_
a +
h
2

¸
¸
¸
.
Usingthefact that
¸
¸
¸
¸
sinθ
θ
¸
¸
¸
¸
< 1andthat |cosθ| ≤ 1, andmakingthesubstitutionh = x −a, weseethat thislast relation
isequivalent to
|sinx −sina| < |x −a|.
Now, toprovethedesiredlimit, let > 0, andtakeδ = . If |x −a| < δ, then
|sinx −sina| < |x −a| < δ = ,
Therefore, aδ wasfoundfor arbitrary, andtheproof iscomplete.
Further Insights and Challenges
Uniquenessof theLimit Provethat afunctionconvergestoat most onelimitingvalue. Inother words, usethe
limit deﬁnitiontoprovethat if lim
x→c
f (x) = L
1
and lim
x→c
f (x) = L
2
, thenL
1
= L
2
.
InExercises31–33, provethestatement usingtheformal limit deﬁnition.
31. TheConstant MultipleLaw[Theorem1, part (ii) inSection2.3, p. 58]
solution Supposethat lim
x→c
f (x) = L. Wewishtoprovethat lim
x→c
af (x) = aL.
Let > 0 begiven. /|a| is also a positivenumber. Since lim
x→c
f (x) = L, weknow thereis a δ > 0 such that
|x −c| < δ forces|f (x) −L| < /|a|. Suppose|x −c| < δ. |af (x) −aL| = |a||f (x) −aL| < |a|(/|a|) = , sothe
ruleisproven.
TheSqueezeTheorem. (Theorem1inSection2.6, p. 77)
June 7, 2011 LTSV SSM Second Pass
82 C HA P T E R 2 LIMITS
33. TheProduct Law[Theorem1, part (iii) inSection2.3, p. 58]. Hint: Usetheidentity
f (x)g(x) −LM = (f (x) −L) g(x) +L(g(x) −M)
solution Before we can prove the Product Law, we need to establish one preliminary result. We are given that
lim
x→c
g(x) = M. Consequently, if weset = 1, thenthedeﬁnitionof alimit guarantees theexistenceof aδ
1
> 0
suchthat whenever 0< |x −c| < δ
1
, |g(x) −M| < 1. Applyingtheinequality|g(x)| −|M| ≤ |g(x) −M|, it follows
that |g(x)| < 1+ |M|. Inother words, becauselim
x→c
g(x) = M, thereexists aδ
1
> 0suchthat |g(x)| < 1+ |M|
whenever 0< |x −c| < δ
1
.
WecannowprovetheProduct Law. Let > 0. Asprovenabove, becauselim
x→c
g(x) = M, thereexistsaδ
1
> 0
suchthat |g(x)| < 1+ |M| whenever 0 < |x − c| < δ
1
. Furthermore, by thedeﬁnitionof alimit, lim
x→c
g(x) = M
impliesthereexistsaδ
2
> 0suchthat |g(x) −M| <

2(1+|L|)
whenever 0< |x −c| < δ
2
. Wehaveincludedthe“1+”
inthedenominator toavoiddivisionbyzeroincaseL = 0. Thereasonfor includingthefactor of 2inthedenominator
will becomeclear shortly. Finally, becauselim
x→c
f (x) = L, thereexists aδ
3
> 0suchthat |f (x) − L| <

2(1+|M|)
whenever 0< |x −c| < δ
3
. Now, let δ = min(δ
1
, δ
2
, δ
3
). Then, for all x satisfying0< |x −c| < δ, wehave
|f (x)g(x) −LM| = |(f (x) −L)g(x) +L(g(x) −M)|
≤ |f (x) −L| |g(x)| +|L| |g(x) −M|
<

2(1+|M|)
(1+|M|) +|L|

2(1+|L|)
<

2
+

2
= .
Hence,
lim
x→c
f (x)g(x) = LM = lim
x→c
f (x) · lim
x→c
g(x).
Let f (x) = 1if x isrational andf (x) = 0if x isirrational. Provethat lim
x→c
f (x) doesnot exist for anyc.
35. Hereisafunctionwithstrangecontinuityproperties:
f (x) =

1
q
if x istherational number p/q in
lowest terms
0 if x isanirrational number
(a) Showthat f (x) isdiscontinuousat c if c isrational. Hint: Thereexist irrational numbersarbitrarilycloseto c.
(b) Showthat f (x) iscontinuousat c if c isirrational. Hint: Let I betheinterval {x : |x −c| < 1}. Showthat for any
Q > 0, I containsat most ﬁnitely many fractionsp/q withq < Q. Concludethat thereisaδ suchthat all fractionsin
{x : |x −c| < δ} haveadenominator larger thanQ.
solution
(a) Let c beanyrational number andsupposethat, inlowest terms, c = p/q, wherep andq areintegers. Toprovethe
discontinuityof f at c, wemust showthereisan > 0suchthat for anyδ > 0thereisanx for which|x −c| < δ, but
that|f (x) −f (c)| > . Let =
1
2q
andδ > 0. Sincethereisatleastoneirrational number betweenanytwodistinctreal
numbers, thereissomeirrational x betweenc andc +δ. Hence, |x −c| < δ, but|f (x) −f (c)| = |0−
1
q
| =
1
q
>
1
2q
= .
(b) Let c beirrational, let > 0 begiven, and let N > 0 beaprimeinteger sufﬁciently largeso that
1
N
< . Let
p
1
q
1
, . . . ,
p
m
q
m
beall rational numbers
p
q
inlowest terms suchthat |
p
q
− c| < 1andq < N. SinceN is ﬁnite, this is a
ﬁnitelist; hence, onenumber
p
i
q
i
inthelist must beclosest toc. Let δ =
1
2
|
p
i
q
i
−c|. Byconstruction, |
p
i
q
i
−c| > δ for all
i = 1. . . m. Therefore, for anyrational number
p
q
suchthat |
p
q
−c| < δ, q > N, so
1
q
<
1
N
< .
Therefore, for anyrational number x suchthat |x −c| < δ, |f (x) −f (c)| < . |f (x) −f (c)| = 0for anyirrational
number x, so|x −c| < δ impliesthat |f (x) −f (c)| < for anynumber x.
CHAPTER REVIEW EXERCISES
1. Thepositionof aparticleat timet (s) iss(t ) =
_
t
2
+1m. Computeitsaveragevelocityover [2, 5] andestimateits
instantaneousvelocityat t = 2.
solution Let s(t ) =
_
t
2
+1. Theaveragevelocityover [2, 5] is
s(5) −s(2)
5−2
=

26−

5
3
≈ 0.954m/s.
Fromthedatainthetablebelow, weestimatethat theinstantaneousvelocityat t = 2isapproximately0.894m/s.
June 7, 2011 LTSV SSM Second Pass
Chapter Review Exercises 83
interval [1.9, 2] [1.99, 2] [1.999, 2] [2, 2.001] [2, 2.01] [2, 2.1]
averageROC 0.889769 0.893978 0.894382 0.894472 0.894873 0.898727
The“wellhead” pricep of natural gasintheUnitedStates(indollarsper 1000ft
3
) ontheﬁrst dayof eachmonth
in2008islistedinthetablebelow.
J F M A M J
6.99 7.55 8.29 8.94 9.81 10.82
J A S O N D
10.62 8.32 7.27 6.36 5.97 5.87
Computetheaveragerateof changeof p (indollarsper1000ft
3
permonth)overthequarterlyperiodsJ anuary–March,
April–J une, andJ uly–September.
3. For awholenumber n, let P(n) bethenumber of partitionsof n, that is, thenumber of waysof writingn asasum
of oneor morewholenumbers. For example, P(4) = 5sincethenumber 4canbepartitionedinﬁvedifferent ways: 4,
3+1, 2+2, 2+1+1, and1+1+1+1. TreatingP(n) asacontinuousfunction, useFigure1toestimatetherateof
changeof P(n) at n = 12.
n
P(n)
14 12 10 8 6 4 2 0
0
40
80
120
160
FIGURE 1 Graphof P(n).
solution Thetangentlinedrawnintheﬁgureappearstopassthroughthepoints(15, 140) and(10.5, 40). Wetherefore
estimatethat therateof changeof P(n) at n = 12is
140−40
15−10.5
=
100
4.5
=
200
9
.
Theaveragevelocity v (m/s) of anoxygenmoleculeintheair at temperatureT (

C) isv = 25.7

273.15+T .
What is theaveragespeedat T = 25

(roomtemperature)? Estimatetherateof changeof averagevelocity with
respect totemperatureat T = 25

. What aretheunitsof thisrate?
InExercises5–10, estimatethelimit numericallytotwodecimal placesor statethat thelimit doesnot exist.
5. lim
x→0
1−cos
3
(x)
x
2
solution Let f (x) =
1−cos
3
x
x
2
. Thedatainthetablebelowsuggeststhat
lim
x→0
1−cos
3
x
x
2
≈ 1.50.
Inconstructingthetable, wetakeadvantageof thefact that f isanevenfunction.
x ±0.001 ±0.01 ±0.1
f (x) 1.500000 1.499912 1.491275
(Theexact valueis
3
2
.)
lim
x→1
x
1/(x−1)
7. lim
x→2
x
x
−4
x
2
−4
solution Let f (x) =
x
x
−4
x
2
−4
. Thedatainthetablebelowsuggeststhat
lim
x→2
x
x
−4
x
2
−4
≈ 1.69.
x 1.9 1.99 1.999 2.001 2.01 2.1
f (x) 1.575461 1.680633 1.691888 1.694408 1.705836 1.828386
(Theexact valueis1+ln2.)
lim
x→2
x −2
2
x
−4
9. lim
x→1
_
7
1−x
7

3
1−x
3
_
solution Let f (x) =
_
7
1−x
7

3
1−x
3
_
. Thedatainthetablebelowsuggeststhat
lim
x→1
_
7
1−x
7

3
1−x
3
_
≈ 2.00.
x 0.9 0.99 0.999 1.001 1.01 1.1
f (x) 2.347483 2.033498 2.003335 1.996668 1.966835 1.685059
(Theexact valueis2.)
June 7, 2011 LTSV SSM Second Pass
84 C HA P T E R 2 LIMITS
lim
x→2
3
x
−9
5
x
−25
InExercises11–50, evaluatethelimit if it exists. If not, determinewhether theone-sidedlimitsexist (ﬁniteor inﬁnite).
11. lim
x→4
(3+x
1/2
)
solution lim
x→4
(3+x
1/2
) = 3+

4= 5.
lim
x→1
5−x
2
4x +7
13. lim
x→−2
4
x
3
solution lim
x→−2
4
x
3
=
4
(−2)
3
= −
1
2
.
lim
x→−1
3x
2
+4x +1
x +1
15. lim
t →9

t −3
t −9
solution lim
t →9

t −3
t −9
= lim
t →9

t −3
(

t −3)(

t +3)
= lim
t →9
1

t +3
=
1

9+3
=
1
6
.
lim
x→3

x +1−2
x −3
17. lim
x→1
x
3
−x
x −1
solution lim
x→1
x
3
−x
x −1
= lim
x→1
x(x −1)(x +1)
x −1
= lim
x→1
x(x +1) = 1(1+1) = 2.
lim
h→0
2(a +h)
2
−2a
2
h
19. lim
t →9
t −6

t −3
solution Becausetheone-sidedlimits
lim
t →9−
t −6

t −3
= −∞ and lim
t →9+
t −6

t −3
= ∞,
arenot equal, thetwo-sidedlimit
lim
t →9
t −6

t −3
doesnot exist.
lim
s→0
1−
_
s
2
+1
s
2
21. lim
x→−1+
1
x +1
solution For x > −1, x +1> 0. Therefore,
lim
x→−1+
1
x +1
= ∞.
lim
y→
1
3
3y
2
+5y −2
6y
2
−5y +1
23. lim
x→1
x
3
−2x
x −1
solution Becausetheone-sidedlimits
lim
x→1−
x
3
−2x
x −1
= ∞ and lim
x→1+
x
3
−2x
x −1
= −∞,
arenot equal, thetwo-sidedlimit
lim
x→1
x
3
−2x
x −1
doesnot exist.
lim
a→b
a
2
−3ab +2b
2
a −b
25. lim
x→0
4
3x
−4
x
4
x
−1
solution
lim
x→0
4
3x
−4
x
4
x
−1
= lim
x→0
4
x
(4
x
−1)(4
x
+1)
4
x
−1
= lim
x→0
4
x
(4
x
+1) = 1· 2= 2.
lim
θ→0
sin5θ
θ
27. lim
x→1.5
[x]
x
solution lim
x→1.5
[x]
x
=
[1.5]
1.5
=
1
1.5
=
2
3
.
June 7, 2011 LTSV SSM Second Pass
Chapter Review Exercises 85
lim
θ→
π
4
secθ 29. lim
z→−3
z +3
z
2
+4z +3
solution
lim
z→−3
z +3
z
2
+4z +3
= lim
z→−3
z +3
(z +3)(z +1)
= lim
z→−3
1
z +1
= −
1
2
.
lim
x→1
x
3
−ax
2
+ax −1
x −1
31. lim
x→b
x
3
−b
3
x −b
solution lim
x→b
x
3
−b
3
x −b
= lim
x→b
(x −b)(x
2
+xb +b
2
)
x −b
= lim
x→b
(x
2
+xb +b
2
) = b
2
+b(b) +b
2
= 3b
2
.
lim
x→0
sin4x
sin3x
33. lim
x→0
_
1
3x

1
x(x +3)
_
solution lim
x→0
_
1
3x

1
x(x +3)
_
= lim
x→0
(x +3) −3
3x(x +3)
= lim
x→0
1
3(x +3)
=
1
3(0+3)
=
1
9
.
lim
θ→
1
4
3
tan(πθ) 35. lim
x→0−
[x]
x
solution For x sufﬁcientlyclosetozerobut negative, [x] = −1. Therefore,
lim
x→0−
[x]
x
= lim
x→0−
−1
x
= ∞.
lim
x→0+
[x]
x
37. lim
θ→
π
2
θ secθ
solution Becausetheone-sidedlimits
lim
θ→
π
2

θ secθ = ∞ and lim
θ→
π
2
+
θ secθ = −∞
arenot equal, thetwo-sidedlimit
lim
θ→
π
2
θ secθ doesnot exist.
lim
y→3
_
sin
π
y
_
−1/2 39. lim
θ→0
cosθ −2
θ
solution Becausetheone-sidedlimits
lim
θ→0−
cosθ −2
θ
= ∞ and lim
θ→0+
cosθ −2
θ
= −∞
arenot equal, thetwo-sidedlimit
lim
θ→0
cosθ −2
θ
doesnot exist.
lim
x→4.3
1
x −[x]
41. lim
x→2−
x −3
x −2
solution For x closeto2but lessthan2, x −3< 0andx −2< 0. Therefore,
lim
x→2−
x −3
x −2
= ∞.
lim
t →0
sin
2
t
t
3
43. lim
x→1+
_
1

x −1

1
_
x
2
−1
_
solution lim
x→1+
_
1

x −1

1
_
x
2
−1
_
= lim
x→1+

x +1−1
_
x
2
−1
= ∞.
lim
t →
π
2

2t (cost −1)
June 7, 2011 LTSV SSM Second Pass
86 C HA P T E R 2 LIMITS
45. lim
x→
π
2
tanx
solution Becausetheone-sidedlimits
lim
x→
π
2

tanx = ∞ and lim
x→
π
2
+
tanx = −∞
arenot equal, thetwo-sidedlimit
lim
x→
π
2
tanx doesnot exist.
lim
t →0
cos
1
t
47. lim
t →0+

t cos
1
t
solution For t > 0,
−1≤ cos
_
1
t
_
≤ 1,
so

t ≤

t cos
_
1
t
_

t .
Because
lim
t →0+

t = lim
t →0+

t = 0,
it followsfromtheSqueezeTheoremthat
lim
t →0+

t cos
_
1
t
_
= 0.
lim
x→5+
x
2
−24
x
2
−25
49. lim
x→0
cosx −1
sinx
solution
lim
x→0
cosx −1
sinx
= lim
x→0
cosx −1
sinx
·
cosx +1
cosx +1
= lim
x→0
−sin
2
x
sinx(cosx +1)
= − lim
x→0
sinx
cosx +1
= −
0
1+1
= 0.
lim
θ→0
tanθ −sinθ
sin
3
θ
51. Findtheleft- andright-handlimits of thefunctionf (x) inFigure2at x = 0, 2, 4. Statewhether f (x) is left- or
right-continuous(or both) at thesepoints.
x
y
1 3 5 2 4
1
2
FIGURE 2
solution Accordingtothegraphof f (x),
lim
x→0−
f (x) = lim
x→0+
f (x) = 1
lim
x→2−
f (x) = lim
x→2+
f (x) = ∞
lim
x→4−
f (x) = −∞
lim
x→4+
f (x) = ∞.
Thefunctionisbothleft- andright-continuousat x = 0andneither left- nor right-continuousat x = 2andx = 4.
Sketchthegraphof afunctionf (x) suchthat
(a) lim
x→2−
f (x) = 1, lim
x→2+
f (x) = 3
(b) lim
x→4
f (x) existsbut doesnot equal f (4).
June 7, 2011 LTSV SSM Second Pass
Chapter Review Exercises 87
53. Graphh(x) anddescribethediscontinuity:
h(x) =
_
2
x
for x ≤ 0
x
−1/2
for x > 0
Ish(x) left- or right-continuous?
solution Thegraphof h(x) isshownbelow.Atx = 0, thefunctionhasaninﬁnitediscontinuitybutisleft-continuous.
2
1
0
−1
−4 −3 −2 −1 1 2 3 4
x
y
Sketchthegraphof afunctiong(x) suchthat
lim
x→−3−
g(x) = ∞, lim
x→−3+
g(x) = −∞, lim
x→4
g(x) = ∞
55. Findthepointsof discontinuityof
g(x) =

cos
_
πx
2
_
for |x| < 1
|x −1| for |x| ≥ 1
Determinethetypeof discontinuityandwhether g(x) isleft- or right-continuous.
solution First notethat cos
_
πx
2
_
iscontinuousfor −1 < x < 1andthat |x − 1| iscontinuousfor x ≤ −1andfor
x ≥ 1. Thus, theonlypointsat whichg(x) might bediscontinuousarex = ±1. At x = 1, wehave
lim
x→1−
g(x) = lim
x→1−
cos
_
πx
2
_
= cos
_
π
2
_
= 0
and
lim
x→1+
g(x) = lim
x→1+
|x −1| = |1−1| = 0,
sog(x) iscontinuousat x = 1. Ontheother hand, at x = −1,
lim
x→−1+
g(x) = lim
x→−1+
cos
_
πx
2
_
= cos
_

π
2
_
= 0
and
lim
x→−1−
g(x) = lim
x→−1−
|x −1| = | −1−1| = 2,
sog(x) hasajumpdiscontinuityat x = −1. Sinceg(−1) = 2, g(x) isleft-continuousat x = −1.
Showthat f (x) = x 2
sinx
iscontinuousonitsdomain.
57. Findaconstant b suchthat h(x) iscontinuousat x = 2, where
h(x) =
_
x +1 for |x| < 2
b −x
2
for |x| ≥ 2
Withthischoiceof b, ﬁndall pointsof discontinuity.
solution Tomakeh(x) continuousatx = 2, wemusthavethetwoone-sidedlimitsasx approaches2beequal. With
lim
x→2−
h(x) = lim
x→2−
(x +1) = 2+1= 3
and
lim
x→2+
h(x) = lim
x→2+
(b −x
2
) = b −4,
itfollowsthatwemustchooseb = 7. Becausex +1iscontinuousfor −2< x < 2and7−x
2
iscontinuousfor x ≤ −2
andfor x ≥ 2, theonlypossiblepoint of discontinuityisx = −2. At x = −2,
lim
x→−2+
h(x) = lim
x→−2+
(x +1) = −2+1= −1
June 7, 2011 LTSV SSM Second Pass
88 C HA P T E R 2 LIMITS
and
lim
x→−2−
h(x) = lim
x→−2−
(7−x
2
) = 7−(−2)
2
= 3,
soh(x) hasajumpdiscontinuityat x = −2.
InExercises58–63, ﬁndthehorizontal asymptotesof thefunctionbycomputingthelimitsat inﬁnity.
f (x) =
9x
2
−4
2x
2
−x
59. f (x) =
x
2
−3x
4
x −1
solution Because
lim
x→∞
x
2
−3x
4
x −1
= lim
x→∞
1/x
2
−3
1/x
3
−1/x
4
= −∞
and
lim
x→−∞
x
2
−3x
4
x −1
= lim
x→−∞
1/x
2
−3
1/x
3
−1/x
4
= ∞,
it followsthat thegraphof y =
x
2
−3x
4
x −1
doesnot haveanyhorizontal asymptotes.
f (u) =
8u −3
_
16u
2
+6
61. f (u) =
2u
2
−1
_
6+u
4
solution Because
lim
u→∞
2u
2
−1
_
6+u
4
= lim
u→∞
2−1/u
2
_
6/u
4
+1
=
2

1
= 2
and
lim
u→−∞
2u
2
−1
_
6+u
4
= lim
u→−∞
2−1/u
2
_
6/u
4
+1
=
2

1
= 2,
it followsthat thegraphof y =
2u
2
−1
_
6+u
4
hasahorizontal asymptoteof y = 2.
f (x) =
3x
2/3
+9x
3/7
7x
4/5
−4x
−1/3
63. f (t ) =
t
1/3
−t
−1/3
(t −t
−1
)
1/3
solution Because
lim
t →∞
t
1/3
−t
−1/3
(t −t
−1
)
1/3
= lim
t →∞
1−t
−2/3
(1−t
−2
)
1/3
=
1
1
1/3
= 1
and
lim
t →−∞
t
1/3
−t
−1/3
(t −t
−1
)
1/3
= lim
t →−∞
1−t
−2/3
(1−t
−2
)
1/3
=
1
1
1/3
= 1,
it followsthat thegraphof y =
t
1/3
−t
−1/3
(t −t
−1
)
1/3
hasahorizontal asymptoteof y = 1.
Calculate(a)–(d), assumingthat
lim
x→3
f (x) = 6, lim
x→3
g(x) = 4
(a) lim
x→3
(f (x) −2g(x)) (b) lim
x→3
x
2
f (x)
(c) lim
x→3
f (x)
g(x) +x
(d) lim
x→3
(2g(x)
3
−g(x)
3/2
)
65. Assumethat thefollowinglimitsexist:
A = lim
x→a
f (x), B = lim
x→a
g(x), L = lim
x→a
f (x)
g(x)
Provethat if L = 1, thenA = B. Hint: Youcannot usetheQuotient Lawif B = 0, soapplytheProduct LawtoL andB
instead.
solution SupposethelimitsA, B, andL all exist andL = 1. Then
B = B · 1= B · L = lim
x→a
g(x) · lim
x→a
f (x)
g(x)
= lim
x→a
g(x)
f (x)
g(x)
= lim
x→a
f (x) = A.
Deﬁneg(t ) = (1+2
1/t
)
−1
for t = 0. Howshouldg(0) bedeﬁnedtomakeg(t ) left-continuousat t = 0?
June 7, 2011 LTSV SSM Second Pass
Chapter Review Exercises 89
67. Inthenotationof Exercise65, giveanexamplewhereL existsbut neither A nor B exists.
solution Suppose
f (x) =
1
(x −a)
3
and g(x) =
1
(x −a)
5
.
Then, neither A nor B exists, but
L = lim
x→a
(x −a)
−3
(x −a)
−5
= lim
x→a
(x −a)
2
= 0.
Trueor false?
(a) If lim
x→3
f (x) exists, then lim
x→3
f (x) = f (3).
(b) If lim
x→0
f (x)
x
= 1, thenf (0) = 0.
(c) If lim
x→−7
f (x) = 8, then lim
x→−7
1
f (x)
=
1
8
.
(d) If lim
x→5+
f (x) = 4and lim
x→5−
f (x) = 8, then lim
x→5
f (x) = 6.
(e) If lim
x→0
f (x)
x
= 1, then lim
x→0
f (x) = 0.
(f) If lim
x→5
f (x) = 2, then lim
x→5
f (x)
3
= 8.
69. Let f (x) = x
_
1
x
_
, where[x] isthegreatest integer function. Showthat for x = 0,
1
x
−1<
_
1
x
_

1
x
ThenusetheSqueezeTheoremtoprovethat
lim
x→0
x
_
1
x
_
= 1
Hint: Treat theone-sidedlimitsseparately.
solution Lety beanyreal number. Fromthedeﬁnitionof thegreatestintegerfunction, itfollowsthaty −1< [y] ≤ y,
withequalityholdingif andonlyif y isaninteger. If x = 0, then
1
x
isareal number, so
1
x
−1<
_
1
x
_

1
x
.
Uponmultiplyingthisinequalitythroughbyx, weﬁnd
1−x < x
_
1
x
_
≤ 1.
Because
lim
x→0
(1−x) = lim
x→0
1= 1,
it followsfromtheSqueezeTheoremthat
lim
x→0
x
_
1
x
_
= 1.
Let r
1
and r
2
be the roots of f (x) = ax
2
− 2x + 20. Observe that f (x) “approaches” the linear function
L(x) = −2x +20asa → 0. Becauser = 10istheuniqueroot of L(x), wemight expect oneof therootsof f (x)
toapproach10asa → 0(Figure3). Provethat therootscanbelabeledsothat lim
a→0
r
1
= 10and lim
a→0
r
2
= ∞.
71. UsetheIVT toprovethat thecurvesy = x
2
andy = cosx intersect.
solution Letf (x) = x
2
−cosx. Notethatanyrootof f (x) correspondstoapointof intersectionbetweenthecurves
y = x
2
and y = cosx. Now, f (x) is continuous over theinterval [0,
π
2
], f (0) = −1 < 0 and f (
π
2
) =
π
2
4
> 0.
Therefore, by theIntermediateValueTheorem, thereexists ac ∈ (0,
π
2
) suchthat f (c) = 0; consequently, thecurves
y = x
2
andy = cosx intersect.
UsetheIVT toprovethat f (x) = x
3

x
2
+2
cosx +2
hasaroot intheinterval [0, 2].
73. UsetheIVT toshowthat 2
−x
2
= x hasasolutionon(0, 1).
solution Let f (x) = 2
−x
2
− x. Observethat f is continuous on[0, 1] withf (0) = 2
0
− 0 = 1 > 0andf (1) =
2
−1
−1< 0. Therefore, theIVT guaranteesthereexistsac ∈ (0, 1) suchthat f (c) = 2
−c
2
−c = 0.
UsetheBisectionMethodtolocateasolutionof x
2
−7= 0totwodecimal places.
75. Giveanexampleof a(discontinuous) functionthat does not satisfy theconclusionof theIVT on[−1, 1].
Thenshowthat thefunction
f (x) =

sin
1
x
x = 0
0 x = 0
satisﬁestheconclusionof theIVT oneveryinterval [−a, a], eventhoughf isdiscontinuousat x = 0.
June 7, 2011 LTSV SSM Second Pass
90 C HA P T E R 2 LIMITS
solution Let g(x) = [x]. Thisfunctionisdiscontinuouson[−1, 1] withg(−1) = −1andg(1) = 1. For all c = 0,
thereisnox suchthatg(x) = c; thus, g(x) doesnotsatisfytheconclusionof theIntermediateValueTheoremon[−1, 1].
Now, let
f (x) =
_
sin
_
1
x
_
for x = 0
0 for x = 0
andlet a > 0. Ontheinterval
x ∈
_
a
2+2πa
,
a
2
_
⊂ [−a, a],
1
x
runsfrom
2
a
to
2
a
+2π, sothesinefunctioncoversonefull periodandclearlytakesoneveryvaluefrom−sina through
sina.
Let f (x) =
1
x +2
.
(a) Showthat
¸
¸
¸f (x) −
1
4
¸
¸
¸ <
|x −2|
12
if |x −2| < 1. Hint: Observethat |4(x +2)| > 12if |x −2| < 1.
(b) Findδ > 0suchthat
¸
¸
¸f (x) −
1
4
¸
¸
¸ < 0.01for |x −2| < δ.
(c) Proverigorouslythat lim
x→2
f (x) =
1
4
.
77. Plot the function f (x) = x
1/3
. Use the zoomfeature to ﬁnd a δ > 0 such that |x
1/3
− 2| < 0.05 for
|x −8| < δ.
solution Thegraphsof y = f (x) = x
1/3
andthehorizontal linesy = 1.95andy = 2.05areshownbelow. From
thisplot, weseethat δ = 0.55guaranteesthat |x
1/3
−2| < 0.05whenever |x −8| < δ.
7 7.5 8 8.5
1.95
1.9
2
2.05
x
y
Usethefact that f (x) = 2
x
is increasingtoﬁndavalueof δ suchthat |2
x
− 8| < 0.001if |x − 2| < δ. Hint:
Findc
1
andc
2
suchthat 7.999< f (c
1
) < f (c
2
) < 8.001.
79. Proverigorouslythat lim
x→−1
(4+8x) = −4.
solution Let > 0andtakeδ = /8. Then, whenever |x −(−1)| = |x +1| < δ,
|f (x) −(−4)| = |4+8x +4| = 8|x +1| < 8δ = .
Proverigorouslythat lim
x→3
(x
2
−x) = 6.
June 8, 2011 LTSV SSM Second Pass
3 DIFFERENTIATION
3.1 Definition of the Derivative
Preliminary Questions
1. Whichof thelinesinFigure10aretangent tothecurve?
A
B
C
D
FIGURE 10
solution LinesB andD aretangent tothecurve.
2. What arethetwowaysof writingthedifferencequotient?
solution Thedifferencequotient maybewritteneither as
f (x) −f (a)
x −a
or as
f (a +h) −f (a)
h
.
3. Finda andh suchthat
f (a +h) −f (a)
h
isequal totheslopeof thesecant linebetween(3, f (3)) and(5, f (5)).
solution Witha = 3andh = 2,
f (a +h) −f (a)
h
isequal totheslopeof thesecantlinebetweenthepoints(3, f (3))
and(5, f (5)) onthegraphof f (x).
4. Whichderivativeisapproximatedby
tan
_
π
4
+0.0001
_
−1
0.0001
?
solution
tan(
π
4
+0.0001) −1
0.0001
isagoodapproximationtothederivativeof thefunctionf (x) = tanx at x =
π
4
.
5. What dothefollowingquantitiesrepresent intermsof thegraphof f (x) = sinx?
(a) sin1.3−sin0.9 (b)
sin1.3−sin0.9
0.4
(c) f

(0.9)
solution Consider thegraphof y = sinx.
(a) Thequantitysin1.3−sin0.9representsthedifferenceinheight betweenthepoints(0.9, sin0.9) and(1.3, sin1.3).
(b) Thequantity
sin1.3−sin0.9
0.4
representstheslopeof thesecantlinebetweenthepoints(0.9, sin0.9) and(1.3, sin1.3)
onthegraph.
(c) Thequantityf

(0.9) representstheslopeof thetangent linetothegraphat x = 0.9.
Exercises
1. Let f (x) = 5x
2
. Showthat f (3+h) = 5h
2
+30h +45. Thenshowthat
f (3+h) −f (3)
h
= 5h +30
andcomputef

(3) bytakingthelimit ash → 0.
91
June 8, 2011 LTSV SSM Second Pass
92 C HA P T E R 3 DIFFERENTIATION
solution Withf (x) = 5x
2
, it followsthat
f (3+h) = 5(3+h)
2
= 5(9+6h +h
2
) = 45+30h +5h
2
.
Usingthisresult, weﬁnd
f (3+h) −f (3)
h
=
45+30h +5h
2
−5· 9
h
=
45+30h +5h
2
−45
h
=
30h +5h
2
h
= 30+5h.
Ash → 0, 30+5h → 30, sof

(3) = 30.
Let f (x) = 2x
2
−3x −5. Showthat thesecant linethrough(2, f (2)) and(2+h, f (2+h)) hasslope2h +5.
Thenusethisformulatocomputetheslopeof:
(a) Thesecant linethrough(2, f (2)) and(3, f (3))
(b) Thetangent lineat x = 2(bytakingalimit)
In Exercises 3–6, compute f

(a) in two ways, using Eq. (1) and Eq. (2).
3. f (x) = x
2
+9x, a = 0
solution Let f (x) = x
2
+9x. Then
f

(0) = lim
h→0
f (0+h) −f (0)
h
= lim
h→0
(0+h)
2
+9(0+h) −0
h
= lim
h→0
9h +h
2
h
= lim
h→0
(9+h) = 9.
Alternately,
f

(0) = lim
x→0
f (x) −f (0)
x −0
= lim
x→0
x
2
+9x −0
x
= lim
x→0
(x +9) = 9.
f (x) = x
2
+9x, a = 2
5. f (x) = 3x
2
+4x +2, a = −1
solution Let f (x) = 3x
2
+4x +2. Then
f

(−1) = lim
h→0
f (−1+h) −f (−1)
h
= lim
h→0
3(−1+h)
2
+4(−1+h) +2−1
h
= lim
h→0
3h
2
−2h
h
= lim
h→0
(3h −2) = −2.
Alternately,
f

(−1) = lim
x→−1
f (x) −f (−1)
x −(−1)
= lim
x→−1
3x
2
+4x +2−1
x +1
= lim
x→−1
(3x +1)(x +1)
x +1
= lim
x→−1
(3x +1) = −2.
f (x) = x
3
, a = 2
In Exercises 7–10, refer to Figure 11.
0.5
1.0
1.5
2.0
2.5
3.0
f (x)
1.0 2.0 3.0 0.5 1.5 2.5
x
y
FIGURE 11
7. Find the slope of the secant line through (2, f (2)) and (2.5, f (2.5)). Is it larger or smaller than f

(2)?
Explain.
solution Fromthegraph, it appears that f (2.5) = 2.5 and f (2) = 2. Thus, theslopeof thesecant linethrough
(2, f (2)) and(2.5, f (2.5)) is
f (2.5) −f (2)
2.5−2
=
2.5−2
2.5−2
= 1.
Fromthegraph, it isalsoclear that thesecant linethrough(2, f (2)) and(2.5, f (2.5)) hasalarger slopethanthetangent
lineat x = 2. Inother words, theslopeof thesecant linethrough(2, f (2)) and(2.5, f (2.5)) islarger thanf

(2).
Estimate
f (2+h) −f (2)
h
for h = −0.5. Whatdoesthisquantityrepresent?Isitlarger or smaller thanf

(2)?
Explain.
9. Estimatef

(1) andf

(2).
solution Fromthegraph, it appearsthat thetangent lineat x = 1wouldbehorizontal. Thus, f

(1) ≈ 0. Thetangent
lineat x = 2appearstopassthroughthepoints(0.5, 0.8) and(2, 2). Thus
f

(2) ≈
2−0.8
2−0.5
= 0.8.
Findavalueof h for which
f (2+h) −f (2)
h
= 0.
June 8, 2011 LTSV SSM Second Pass
S E C T I ON 3.1 Definition of the Derivative 93
In Exercises 11–14, refer to Figure 12.
1
2
3
5
4
1 2 3 4 5 6 7 8 9
x
y
FIGURE 12 Graphof f (x).
11. Determinef

(a) for a = 1, 2, 4, 7.
solution Remember that thevalueof thederivativeof f at x = a canbeinterpretedastheslopeof thelinetangent
tothegraphof y = f (x) at x = a. FromFigure12, weseethat thegraphof y = f (x) isahorizontal line(that is, aline
withzeroslope) ontheinterval 0 ≤ x ≤ 3. Accordingly, f

(1) = f

(2) = 0. Ontheinterval 3 ≤ x ≤ 5, thegraphof
y = f (x) isalineof slope
1
2
; thus, f

(4) =
1
2
. Finally, thelinetangent tothegraphof y = f (x) at x = 7ishorizontal,
sof

(7) = 0.
For whichvaluesof x isf

(x) < 0?
13. Whichislarger, f

(5.5) or f

(6.5)?
solution Thelinetangent tothegraphof y = f (x) at x = 5.5hasalarger slopethanthelinetangent tothegraphof
y = f (x) at x = 6.5. Therefore, f

(5.5) islarger thanf

(6.5).
Showthat f

(3) doesnot exist.
In Exercises 15–18, use the limit deﬁnition to calculate the derivative of the linear function.
15. f (x) = 7x −9
solution
lim
h→0
f (a +h) −f (a)
h
= lim
h→0
7(a +h) −9−(7a −9)
h
= lim
h→0
7= 7.
f (x) = 12
17. g(t ) = 8−3t
solution
lim
h→0
g(a +h) −g(a)
h
= lim
h→0
8−3(a +h) −(8−3a)
h
= lim
h→0
−3h
h
= lim
h→0
(−3) = −3.
k(z) = 14z +12
19. Findanequationof thetangent lineat x = 3, assumingthat f (3) = 5andf

(3) = 2?
solution Bydeﬁnition, theequationof thetangent linetothegraphof f (x) at x = 3isy = f (3) +f

(3)(x −3) =
5+2(x −3) = 2x −1.
Findf (3) andf

(3), assumingthat thetangent linetoy = f (x) at a = 3hasequationy = 5x +2.
21. Describethetangent lineat anarbitrarypoint onthe“curve” y = 2x +8.
solution Sincey = 2x +8representsastraight line, thetangent lineat anypoint isthelineitself, y = 2x +8.
Supposethat f (2+h) −f (2) = 3h
2
+5h. Calculate:
(a) Theslopeof thesecant linethrough(2, f (2)) and(6, f (6))
(b) f

(2)
23. Let f (x) =
1
x
. Does f (−2+ h) equal
1
−2+h
or
1
−2
+
1
h
? Compute the difference quotient at a = −2 with
h = 0.5.
solution Let f (x) =
1
x
. Then
f (−2+h) =
1
−2+h
.
Witha = −2andh = 0.5, thedifferencequotient is
f (a +h) −f (a)
h
=
f (−1.5) −f (−2)
0.5
=
1
−1.5

1
−2
0.5
= −
1
3
.
Let f (x) =

x. Does f (5+ h) equal

5+h or

5+

h? Computethedifferencequotient at a = 5with
h = 1.
25. Let f (x) = 1/

x. Computef

(5) byshowingthat
f (5+h) −f (5)
h
= −
1

5

5+h(

5+h +

5)
June 8, 2011 LTSV SSM Second Pass
94 C HA P T E R 3 DIFFERENTIATION
solution Let f (x) = 1/

x. Then
f (5+h) −f (5)
h
=
1

5+h

1

5
h
=

5−

5+h
h

5

5+h
=

5−

5+h
h

5

5+h
_

5+

5+h

5+

5+h
_
=
5−(5+h)
h

5

5+h(

5+h +

5)
= −
1

5

5+h(

5+h +

5)
.
Thus,
f

(5) = lim
h→0
f (5+h) −f (5)
h
= lim
h→0

1

5

5+h(

5+h +

5)
= −
1

5

5(

5+

5)
= −
1
10

5
.
Findanequationof thetangent linetothegraphof f (x) = 1/

x at x = 9.
In Exercises 27–44, use the limit deﬁnition to compute f

(a) and ﬁnd an equation of the tangent line.
27. f (x) = 2x
2
+10x, a = 3
solution Let f (x) = 2x
2
+10x. Then
f

(3) = lim
h→0
f (3+h) −f (3)
h
= lim
h→0
2(3+h)
2
+10(3+h) −48
h
= lim
h→0
18+12h +2h
2
+30+10h −48
h
= lim
h→0
(22+2h) = 22.
At a = 3, thetangent lineis
y = f

(3)(x −3) +f (3) = 22(x −3) +48= 22x −18.
f (x) = 4−x
2
, a = −1
29. f (t ) = t −2t
2
, a = 3
solution Let f (t ) = t −2t
2
. Then
f

(3) = lim
h→0
f (3+h) −f (3)
h
= lim
h→0
(3+h) −2(3+h)
2
−(−15)
h
= lim
h→0
3+h −18−12h −2h
2
+15
h
= lim
h→0
(−11−2h) = −11.
At a = 3, thetangent lineis
y = f

(3)(t −3) +f (3) = −11(t −3) −15= −11t +18.
f (x) = 8x
3
, a = 1
31. f (x) = x
3
+x, a = 0
solution Let f (x) = x
3
+x. Then
f

(0) = lim
h→0
f (h) −f (0)
h
= lim
h→0
h
3
+h −0
h
= lim
h→0
(h
2
+1) = 1.
At a = 0, thetangent lineis
y = f

(0)(x −0) +f (0) = x.
f (t ) = 2t
3
+4t , a = 4
33. f (x) = x
−1
, a = 8
solution Let f (x) = x
−1
. Then
f

(8) = lim
h→0
f (8+h) −f (8)
h
= lim
h→0
1
8+h

_
1
8
_
h
= lim
h→0
8−8−h
8(8+h)
h
= lim
h→0
−h
(64+8h)h
= −
1
64
June 8, 2011 LTSV SSM Second Pass
S E C T I ON 3.1 Definition of the Derivative 95
Thetangent at a = 8is
y = f

(8)(x −8) +f (8) = −
1
64
(x −8) +
1
8
= −
1
64
x +
1
4
.
f (x) = x +x
−1
, a = 4
35. f (x) =
1
x +3
, a = −2
solution Let f (x) =
1
x+3
. Then
f

(−2) = lim
h→0
f (−2+h) −f (−2)
h
= lim
h→0
1
−2+h+3
−1
h
= lim
h→0
1
1+h
−1
h
= lim
h→0
−h
h(1+h)
= lim
h→0
−1
1+h
= −1.
Thetangent lineat a = −2is
y = f

(−2)(x +2) +f (−2) = −1(x +2) +1= −x −1.
f (t ) =
2
1−t
, a = −1
37. f (x) =

x +4, a = 1
solution Let f (x) =

x +4. Then
f

(1) = lim
h→0
f (1+h) −f (1)
h
= lim
h→0

h +5−

5
h
= lim
h→0

h +5−

5
h
·

h +5+

5

h +5+

5
= lim
h→0
h
h(

h +5+

5)
= lim
h→0
1

h +5+

5
=
1
2

5
.
Thetangent lineat a = 1is
y = f

(1)(x −1) +f (1) =
1
2

5
(x −1) +

5=
1
2

5
x +
9
2

5
.
f (t ) =

3t +5, a = −1
39. f (x) =
1

x
, a = 4
solution Let f (x) =
1

x
. Then
f

(4) = lim
h→0
f (4+h) −f (4)
h
= lim
h→0
1

4+h

1
2
h
= lim
h→0
2−

4+h
2

4+h
·
2+

4+h
2+

4+h
h
= lim
h→0
4−4−h
4

4+h+2(4+h)
h
= lim
h→0
−1
4

4+h +2(4+h)
= −
1
16
.
At a = 4thetangent lineis
y = f

(4)(x −4) +f (4) = −
1
16
(x −4) +
1
2
= −
1
16
x +
3
4
.
f (x) =
1

2x +1
, a = 4
41. f (t ) =
_
t
2
+1, a = 3
solution Let f (t ) =
_
t
2
+1. Then
f

(3) = lim
h→0
f (3+h) −f (3)
h
= lim
h→0
_
10+6h +h
2

10
h
= lim
h→0
_
10+6h +h
2

10
h
·
_
10+6h +h
2
+

10
_
10+6h +h
2
+

10
= lim
h→0
6h +h
2
h(
_
10+6h +h
2
+

10)
= lim
h→0
6+h
_
10+6h +h
2
+

10
=
3

10
.
Thetangent lineat a = 3is
y = f

(3)(t −3) +f (3) =
3

10
(t −3) +

10=
3

10
t +
1

10
.
June 8, 2011 LTSV SSM Second Pass
96 C HA P T E R 3 DIFFERENTIATION
f (x) = x
−2
, a = −1
43. f (x) =
1
x
2
+1
, a = 0
solution Let f (x) =
1
x
2
+1
. Then
f

(0) = lim
h→0
f (0+h) −f (0)
h
= lim
h→0
1
(0+h)
2
+1
−1
h
= lim
h→0
−h
2
h
2
+1
h
= lim
h→0
−h
h
2
+1
= 0.
Thetangent lineat a = 0is
y = f (0) +f

(0)(x −0) = 1+0(x −1) = 1.
f (t ) = t
−3
, a = 1
45. Figure13displaysdatacollectedbythebiologistJ ulianHuxley(1887–1975) ontheaverageantler weightW of male
reddeer asafunctionof aget . Estimatethederivativeatt = 4. For whichvaluesof t istheslopeof thetangentlineequal
tozero? For whichvaluesisit negative?
2 4 0 6 8 10 12 14
Age(years)
Antler
Weight
(kg)
0
1
2
3
4
5
6
7
8
FIGURE 13
solution LetW(t ) denotetheantlerweightasafunctionof age. The“tangentline”sketchedintheﬁgurebelowpasses
throughthepoints(1, 1) and(6, 5.5). Therefore
W

(4) ≈
5.5−1
6−1
= 0.9kg/year.
If theslopeof thetangentiszero, thetangentlineishorizontal. Thisappearstohappenatroughlyt = 10andatt = 11.6.
Theslopeof thetangent lineisnegativewhentheheight of thegraphdecreasesaswemovetotheright. For thegraphin
Figure13, thisoccursfor 10< t < 11.6.
2 4 0 6 8 10 12 14
0
1
2
3
4
5
6
7
8
y
x
Figure14(A) shows thegraph of f (x) =

x. Theclose-up in Figure14(B) shows that thegraph is nearly a
straight linenear x = 16. Estimatetheslopeof thislineandtakeit asanestimatefor f

(16). Thencomputef

(16)
andcomparewithyour estimate.
47. Let f (x) =
4
1+2
x
.
(a) Plot f (x) over [−2, 2]. Thenzoominnear x = 0until thegraphappearsstraight, andestimatetheslopef

(0).
(b) Use(a) toﬁndanapproximateequationtothetangent lineat x = 0. Plot thislineandf (x) onthesameset of axes.
solution
(a) Theﬁgurebelowattheleftshowsthegraphof f (x) =
4
1+2
x
over [−2, 2]. Theﬁgurebelowattherightisaclose-up
near x = 0. Fromtheclose-up, weseethat thegraphisnearly straight andpassesthroughthepoints(−0.22, 2.15) and
(0.22, 1.85). Wethereforeestimate
f

(0) ≈
1.85−2.15
0.22−(−0.22)
=
−0.3
0.44
= −0.68
y
x
−2 −1 1 2
y
x
−0.2 −0.1 0.1 0.2
0.5
1.8
2.0
2.2
2.4
1
1.5
2
3
2.5
June 8, 2011 LTSV SSM Second Pass
S E C T I ON 3.1 Definition of the Derivative 97
(b) Usingtheestimatefor f

(0) obtainedinpart (a), theapproximateequationof thetangent lineis
y = f

(0)(x −0) +f (0) = −0.68x +2.
Theﬁgurebelowshowsthegraphof f (x) andtheapproximatetangent line.
y
x
−2 −1 1 2
0.5
1
1.5
2
3
2.5
Let f (x) = cotx. Estimatef

_
π
2
_
graphicallybyzoominginonaplot of f (x) near x =
π
2
.
49. Determinetheintervalsalongthex-axisonwhichthederivativeinFigure15ispositive.
1.0 1.5 2.0 2.5 3.0 3.5 4.0 0.5
1.0
0.5
1.5
2.0
2.5
3.0
3.5
4.0
x
y
FIGURE 15
solution Thederivative(thatis, theslopeof thetangentline) ispositivewhentheheightof thegraphincreasesaswe
movetotheright. FromFigure15, thisappearstobetruefor 1< x < 2.5andfor x > 3.5.
Sketchthegraphof f (x) = sinx on[0, π] andguessthevalueof f

_
π
2
_
. Thencalculatethedifferencequotient
at x =
π
2
for twosmall positiveandnegativevaluesof h. Arethesecalculationsconsistent withyour guess?
In Exercises 51–56, each limit represents a derivative f

(a). Find f (x) and a.
51. lim
h→0
(5+h)
3
−125
h
solution Thedifferencequotient
(5+h)
3
−125
h
hastheform
f (a +h) −f (a)
h
wheref (x) = x
3
anda = 5.
lim
x→5
x
3
−125
x −5
53. lim
h→0
sin
_
π
6
+h
_
−0.5
h
solution Thedifferencequotient
sin(
π
6
+h) −0.5
h
hastheform
f (a +h) −f (a)
h
wheref (x) = sinx anda =
π
6
.
lim
x→
1
4
x
−1
−4
x −
1
4
55. lim
h→0
5
2+h
−25
h
solution Thedifferencequotient
5
(2+h)
−25
h
hastheform
f (a +h) −f (a)
h
wheref (x) = 5
x
anda = 2.
lim
h→0
5
h
−1
h
57. Applythemethodof Example6tof (x) = sinx todeterminef

_
π
4
_
accuratelytofour decimal places.
solution Weknowthat
f

(π/4) = lim
h→0
f (π/4+h) −f (π/4)
h
= lim
h→0
sin(π/4+h) −

2/2
h
.
Creatingatableof valuesof h closetozero:
h −0.001 −0.0001 −0.00001 0.00001 0.0001 0.001
sin(
π
4
+h) −(

2/2)
h
0.7074602 0.7071421 0.7071103 0.7071033 0.7070714 0.7067531
Accurateuptofour decimal places, f

(
π
4
) ≈ 0.7071.
Applythemethodof Example6tof (x) = cosx todeterminef

_
π
5
_
accuratelytofour decimal places. Usea
graphof f (x) toexplainhowthemethodworksinthiscase.
June 8, 2011 LTSV SSM Second Pass
98 C HA P T E R 3 DIFFERENTIATION
59. For eachgraphinFigure16, determinewhether f

(1) islarger or smaller thantheslopeof thesecant line
betweenx = 1andx = 1+h for h > 0. Explain.
1 1
(A) (B)
y
x
y
x
y = f (x)
y = f (x)
FIGURE 16
solution
• Oncurve(A),f

(1) islarger than
f (1+h) −f (1)
h
;
thecurveisbendingdownwards, sothat thesecant linetotheright isat alower anglethanthetangent line. Wesay
suchacurveisconcavedown, andthat itsderivativeisdecreasing.
• Oncurve(B), f

(1) issmaller than
f (1+h) −f (1)
h
;
thecurveisbendingupwards, sothat thesecant linetotheright isat asteeper anglethanthetangent line. Wesay
suchacurveisconcaveup, andthat itsderivativeisincreasing.
Refer tothegraphof f (x) = 2
x
inFigure17.
(a) Explaingraphicallywhy, for h > 0,
f (−h) −f (0)
−h
≤ f

(0) ≤
f (h) −f (0)
h
(b) Use(a) toshowthat 0.69314≤ f

(0) ≤ 0.69315.
(c) Similarly, computef

(x) tofour decimal placesfor x = 1, 2, 3, 4.
(d) Nowcomputetheratiosf

(x)/f

(0) for x = 1, 2, 3, 4. Canyouguessanapproximateformulafor f

(x)?
61. Sketchthegraphof f (x) = x
5/2
on[0, 6].
(a) Usethesketchtojustifytheinequalitiesfor h > 0:
f (4) −f (4−h)
h
≤ f

(4) ≤
f (4+h) −f (4)
h
(b) Use(a) tocomputef

(4) tofour decimal places.
(c) Useagraphingutilitytoplot f (x) andthetangent lineat x = 4, usingyour estimatefor f

(4).
solution
(a) Theslopeof thesecant linebetweenpoints(4, f (4)) and(4+h, f (4+h)) is
f (4+h) −f (4)
h
.
x
5/2
is asmoothcurveincreasingat afaster rateas x → ∞. Therefore, if h > 0, thentheslopeof thesecant lineis
greater thantheslopeof thetangent lineat f (4), whichhappenstobef

(4). Likewise, if h < 0, theslopeof thesecant
lineislessthantheslopeof thetangent lineat f (4), whichhappenstobef

(4).
(b) Weknowthat
f

(4) = lim
h→0
f (4+h) −f (4)
h
= lim
h→0
(4+h)
5/2
−32
h
.
Creatingatablewithvaluesof h closetozero:
h −0.0001 −0.00001 0.00001 0.0001
(4+h)
5/2
−32
h
19.999625 19.99999 20.0000 20.0000375
Thus, f

(4) ≈ 20.0000.
(c) Usingtheestimatefor f

(4) obtainedinpart (b), theequationof thelinetangent tof (x) = x
5/2
at x = 4is
y = f

(4)(x −4) +f (4) = 20(x −4) +32= 20x −48.
June 8, 2011 LTSV SSM Second Pass
S E C T I ON 3.1 Definition of the Derivative 99
y
x
1 2 3 4 5 6
−20
−40
−60
20
40
60
80
Verify that P =
_
1,
1
2
_
lies onthegraphs of bothf (x) = 1/(1+x
2
) andL(x) =
1
2
+ m(x − 1) for every
slopem. Plotf (x) andL(x) onthesameaxesfor several valuesof muntil youﬁndavalueof mfor whichy = L(x)
appearstangent tothegraphof f (x). What isyour estimatefor f

(1)?
63. Useaplot of f (x) = x
x
toestimatethevaluec suchthat f

(c) = 0. Findc tosufﬁcient accuracysothat
¸
¸
¸
¸
f (c +h) −f (c)
h
¸
¸
¸
¸
≤ 0.006 for h = ±0.001
solution Hereisagraphof f (x) = x
x
over theinterval [0, 1.5].
0.2 0.4 0.6 0.8 1 1.2 1.4
0.5
1
1.5
2
y
x
Thegraphshowsonelocationwithahorizontal tangentline. Theﬁgurebelowattheleftshowsthegraphof f (x) together
withthehorizontal linesy = 0.6, y = 0.7andy = 0.8. Theliney = 0.7isvery closetobeingtangent tothegraphof
f (x). Theﬁgurebelowattherightreﬁnesthisestimatebygraphingf (x) andy = 0.69onthesamesetof axes. Thepoint
of tangencyhasanx-coordinateof roughly0.37, soc ≈ 0.37.
0.2 0.4 0.6 0.8 1 1.2 1.4
0.5
1
1.5
2
y
x
0.2 0.4 0.6 0.8 1 1.2 1.4
0.5
1
1.5
2
y
x
Wenotethat
¸
¸
¸
¸
f (0.37+0.001) −f (0.37)
0.001
¸
¸
¸
¸
≈ 0.00491< 0.006
and
¸
¸
¸
¸
f (0.37−0.001) −f (0.37)
0.001
¸
¸
¸
¸
≈ 0.00304< 0.006,
sowehavedeterminedc tothedesiredaccuracy.
Plot f (x) = x
x
andy = 2x +a onthesameset of axesfor several valuesof a until thelinebecomestangent
tothegraph. Thenestimatethevaluec suchthat f

(c) = 2.
In Exercises 65–71, estimate derivatives using the symmetric difference quotient (SDQ), deﬁned as the average of the
difference quotients at h and −h:
1
2
_
f (a +h) −f (a)
h
+
f (a −h) −f (a)
−h
_
=
f (a +h) −f (a −h)
2h
4
The SDQ usually gives a better approximation to the derivative than the difference quotient.
65. Thevapor pressureof water at temperatureT (inkelvins) istheatmosphericpressureP at whichnonet evaporation
takesplace. UsethefollowingtabletoestimateP

(T ) for T = 303, 313, 323, 333, 343bycomputingtheSDQgivenby
Eq. (4) withh = 10.
T (K) 293 303 313 323 333 343 353
P (atm) 0.0278 0.0482 0.0808 0.1311 0.2067 0.3173 0.4754
June 8, 2011 LTSV SSM Second Pass
100 C HA P T E R 3 DIFFERENTIATION
solution Usingequation(4),
P

(303) ≈
P(313) −P(293)
20
=
0.0808−0.0278
20
= 0.00265atm/K;
P

(313) ≈
P(323) −P(303)
20
=
0.1311−0.0482
20
= 0.004145atm/K;
P

(323) ≈
P(333) −P(313)
20
=
0.2067−0.0808
20
= 0.006295atm/K;
P

(333) ≈
P(343) −P(323)
20
=
0.3173−0.1311
20
= 0.00931atm/K;
P

(343) ≈
P(353) −P(333)
20
=
0.4754−0.2067
20
= 0.013435atm/K
UsetheSDQ withh = 1year to estimateP

(T ) intheyears 2000, 2002, 2004, 2006, whereP(T ) is theU.S.
ethanol production(Figure18). Expressyour answer inthecorrect units.
In Exercises 67 and 68, trafﬁc speed S along a certain road (in km/h) varies as a function of trafﬁc density q (number of
cars per km of road). Use the following data to answer the questions:
q (density) 60 70 80 90 100
S (speed) 72.5 67.5 63.5 60 56
67. EstimateS

(80).
solution Let S(q) bethefunctiondeterminingS givenq. Usingequation(4) withh = 10,
S

(80) ≈
S(90) −S(70)
20
=
60−67.5
20
= −0.375;
withh = 20,
S

(80) ≈
S(100) −S(60)
40
=
56−72.5
40
= −0.4125;
Themeanof thesetwosymmetricdifferencequotientsis−0.39375kph·km/car.
Explainwhy V = qS, calledtrafﬁc volume, is equal tothenumber of cars passingapoint per hour. Usethe
datatoestimateV

(80).
Exercises 69–71: The current (in amperes) at time t (in seconds) ﬂowing in the circuit in Figure 19 is given by Kirchhoff ’s
Law:
i(t ) = Cv

(t ) +R
−1
v(t )
where v(t ) is the voltage (in volts), C the capacitance (in farads), and R the resistance (in ohms, ).
+

v
R
i
C
FIGURE 19
69. Calculatethecurrent at t = 3if
v(t ) = 0.5t +4V
whereC = 0.01F andR = 100.
solution Since v(t ) is a line with slope 0.5, v

(t ) = 0.5 volts/s for all t . Fromthe formula, i(3) = Cv

(3) +
(1/R)v(3) = 0.01(0.5) +(1/100)(5.5) = 0.005+0.055= 0.06amperes.
Usethefollowingdatatoestimatev

(10) (byanSDQ). Thenestimatei(10), assumingC = 0.03andR = 1000.
t 9.8 9.9 10 10.1 10.2
v(t ) 256.52 257.32 258.11 258.9 259.69
71. Assumethat R = 200 but C is unknown. Usethefollowingdatato estimatev

(4) (by anSDQ) anddeducean
approximatevaluefor thecapacitanceC.
t 3.8 3.9 4 4.1 4.2
v(t ) 388.8 404.2 420 436.2 452.8
i(t ) 32.34 33.22 34.1 34.98 35.86
June 8, 2011 LTSV SSM Second Pass
S E C T I ON 3.2 The Derivative as a Function 101
solution Solvingi(4) = Cv

(4) +(1/R)v(4) for C yields
C =
i(4) −(1/R)v(4)
v

(4)
=
34.1−
420
200
v

(4)
.
TocomputeC, weﬁrst approximatev

(4). Takingh = 0.1, weﬁnd
v

(4) ≈
v(4.1) −v(3.9)
0.2
=
436.2−404.2
0.2
= 160.
Pluggingthisintotheequationaboveyields
C ≈
34.1−2.1
160
= 0.2farads.
Further Insights and Challenges
TheSDQusuallyapproximatesthederivativemuchmorecloselythandoestheordinarydifferencequotient. Let
f (x) = 2
x
anda = 0. ComputetheSDQwithh = 0.001andtheordinary differencequotientswithh = ±0.001.
Comparewiththeactual value, whichisf

(0) = ln2.
73. Explainhowthesymmetricdifferencequotient deﬁnedbyEq. (4) canbeinterpretedastheslopeof asecant line.
solution Thesymmetricdifferencequotient
f (a +h) −f (a −h)
2h
is theslopeof thesecant lineconnectingthepoints (a − h, f (a − h)) and(a + h, f (a + h)) onthegraphof f ; the
differenceinthefunctionvaluesisdividedbythedifferenceinthex-values.
Whichof thetwofunctionsinFigure20satisﬁestheinequality
f (a +h) −f (a −h)
2h

f (a +h) −f (a)
h
for h > 0? Explainintermsof secant lines.
75. Showthat if f (x) is aquadratic polynomial, then theSDQ at x = a (for any h = 0) is equal to f

(a).
Explainthegraphical meaningof thisresult.
solution Let f (x) = px
2
+qx +r beaquadraticpolynomial. WecomputetheSDQat x = a.
f (a +h) −f (a −h)
2h
=
p(a +h)
2
+q(a +h) +r −(p(a −h)
2
+q(a −h) +r)
2h
=
pa
2
+2pah +ph
2
+qa +qh +r −pa
2
+2pah −ph
2
−qa +qh −r
2h
=
4pah +2qh
2h
=
2h(2pa +q)
2h
= 2pa +q
Sincethisdoesn’tdependonh, thelimit, whichisequal tof

(a), isalso2pa +q. Graphically, thisresulttellsusthatthe
secant linetoaparabolapassingthroughpointschosensymmetricallyabout x = a isalwaysparallel tothetangent line
at x = a.
Let f (x) = x
−2
. Computef

(1) bytakingthelimit of theSDQs(witha = 1) ash → 0.
3.2 The Derivative as a Function
Preliminary Questions
1. What istheslopeof thetangent linethroughthepoint (2, f (2)) if f

(x) = x
3
?
solution Theslopeof thetangentlinethroughthepoint(2, f (2)) isgivenbyf

(2). Sincef

(x) = x
3
, itfollowsthat
f

(2) = 2
3
= 8.
2. Evaluate(f −g)

(1) and(3f +2g)

(1) assumingthat f

(1) = 3andg

(1) = 5.
solution (f −g)

(1) = f

(1) −g

(1) = 3−5= −2and(3f +2g)

(1) = 3f

(1) +2g

(1) = 3(3) +2(5) = 19.
3. Towhichof thefollowingdoesthePower Ruleapply?
(a) f (x) = x
2
(b) f (x) = 2
π
(c) f (x) = x
π
(d) f (x) = π
x
(e) f (x) = x
x
(f) f (x) = x
−4/5
solution
(a) Yes. x
2
isapower function, sothePower Rulecanbeapplied.
(b) Yes. 2
π
isaconstant function, sothePower Rulecanbeapplied.
(c) Yes. x
π
isapower function, sothePower Rulecanbeapplied.
(d) No. π
x
isanexponential function(thebaseisconstant whiletheexponent isavariable), sothePower Ruledoesnot
apply.
June 8, 2011 LTSV SSM Second Pass
102 C HA P T E R 3 DIFFERENTIATION
(e) No. x
x
isnotapower functionbecauseboththebaseandtheexponentarevariable, sothePower Ruledoesnotapply.
(f) Yes. x
−4/5
isapower function, sothePower Rulecanbeapplied.
4. Choose(a) or (b). Thederivativedoesnot exist if thetangent lineis: (a) horizontal (b) vertical.
solution Thederivativedoesnot exist when: (b) thetangent lineisvertical. At ahorizontal tangent, thederivativeis
zero.
5. If f (x) isdifferentiableat x = c, isf (x) necessarilycontinuousat x = c? Dothereexist continuousfunctionsthat
arenot differentiable?
solution ByTheorem4, if f isdifferentiableat x = c, thenit iscontinuousat x = c. Theconversedoesnot hold,
however. For example, f (x) = |x| is continuous at x = 0but is not differentiabletheresincetheslopes fromtheleft
equal −1whilethosefromtheright equal 1.
Exercises
In Exercises 1–6, compute f

(x) using the limit deﬁnition.
1. f (x) = 3x −7
solution Let f (x) = 3x −7. Then,
f

(x) = lim
h→0
f (x +h) −f (x)
h
= lim
h→0
3(x +h) −7−(3x −7)
h
= lim
h→0
3h
h
= 3.
f (x) = x
2
+3x
3. f (x) = x
3
solution Let f (x) = x
3
. Then,
f

(x) = lim
h→0
f (x +h) −f (x)
h
= lim
h→0
(x +h)
3
−x
3
h
= lim
h→0
x
3
+3x
2
h +3xh
2
+h
3
−x
3
h
= lim
h→0
3x
2
h +3xh
2
+h
3
h
= lim
h→0
(3x
2
+3xh +h
2
) = 3x
2
.
f (x) = 1−x
−1
5. f (x) = x −

x
solution Let f (x) = x −

x. Then,
f

(x) = lim
h→0
f (x +h) −f (x)
h
= lim
h→0
x +h −

x +h −(x −

x)
h
= 1− lim
h→0

x +h −

x
h
·
_

x +h +

x

x +h +

x
_
= 1− lim
h→0
(x +h) −x
h(

x +h +

x)
= 1− lim
h→0
1

x +h +

x
= 1−
1
2

x
.
f (x) = x
−1/2
In Exercises 7–14, use the Power Rule to compute the derivative.
7.
d
dx
x
4
¸
¸
¸
¸
x=−2
solution
d
dx
_
x
4
_
= 4x
3
so
d
dx
x
4
¸
¸
¸
¸
x=−2
= 4(−2)
3
= −32.
d
dt
t
−3
¸
¸
¸
¸
t =4
9.
d
dt
t
2/3
¸
¸
¸
¸
t =8
solution
d
dt
_
t
2/3
_
=
2
3
t
−1/3
so
d
dt
t
2/3
¸
¸
¸
¸
t =8
=
2
3
(8)
−1/3
=
1
3
.
d
dt
t
−2/5
¸
¸
¸
¸
t =1
11.
d
dx
x
0.35
solution
d
dx
_
x
0.35
_
= 0.35(x
0.35−1
) = 0.35x
−0.65
.
d
dx
x
14/3
13.
d
dt
t

17
solution
d
dt
_
t

17
_
=

17t

17−1
d
dt
t
−π
2
June 8, 2011 LTSV SSM Second Pass
S E C T I ON 3.2 The Derivative as a Function 103
In Exercises 15–18, compute f

(x) and ﬁnd an equation of the tangent line to the graph at x = a.
15. f (x) = x
4
, a = 2
solution Let f (x) = x
4
. Then, by thePower Rule, f

(x) = 4x
3
. Theequationof thetangent lineto thegraphof
f (x) at x = 2is
y = f

(2)(x −2) +f (2) = 32(x −2) +16= 32x −48.
f (x) = x
−2
, a = 5
17. f (x) = 5x −32

x, a = 4
solution Letf (x) = 5x −32x
1/2
. Thenf

(x) = 5−16x
−1/2
. Inparticular, f

(4) = −3. Thetangentlineatx = 4
is
y = f

(4)(x −4) +f (4) = −3(x −4) −44= −3x −32.
f (x) =
3

x, a = 8
19. Findanequationof thetangent linetoy =
1
x
at x = 9.
solution Let f (x) =
1
x
. Thenf (9) =
1
9
, f

(x) = −
1
x
2
, andf

(9) = −
1
81
. Theequationof thetangent lineis
y −
1
9
= −
1
81
(x −9), or y = −
1
81
x +
2
9
Findapoint onthegraphof y =

x wherethetangent linehasslope10.
In Exercises 21–32, calculate the derivative.
21. f (x) = 2x
3
−3x
2
+5
solution
d
dx
_
2x
3
−3x
2
+5
_
= 6x
2
−6x.
f (x) = 2x
3
−3x
2
+2x
23. f (x) = 4x
5/3
−3x
−2
−12
solution
d
dx
_
4x
5/3
−3x
−2
−12
_
=
20
3
x
2/3
+6x
−3
.
f (x) = x
5/4
+4x
−3/2
+11x
25. g(z) = 7z
−5/14
+z
−5
+9
solution
d
dz
_
7z
−5/14
+z
−5
+9
_
= −
5
2
z
−19/14
−5z
−6
.
h(t ) = 6

t +
1

t
27. f (s) =
4

s +
3

s
solution f (s) =
4

s +
3

s = s
1/4
+s
1/3
. Inthisform, wecanapplytheSumandPower Rules.
d
ds
_
s
1/4
+s
1/3
_
=
1
4
(s
(1/4)−1
) +
1
3
(s
(1/3)−1
) =
1
4
s
−3/4
+
1
3
s
−2/3
.
W(y) = 6y
4
+7y
2/3
29. g(x) = π
2
solution Becauseπ
2
isaconstant,
d
dx
π
2
= 0.
f (x) = x
π 31. h(t ) =

2t

2
solution
d
dt

2t

2
= 2t

2−1
.
R(z) =
z
5/3
−4z
3/2
z
Hint: simplify.
In Exercises 33–36, calculate the derivative by expanding or simplifying the function.
33. P(s) = (4s −3)
2
solution P(s) = (4s −3)
2
= 16s
2
−24s +9. Thus,
dP
ds
= 32s −24.
Q(r) = (1−2r)(3r +5)
35. g(x) =
x
2
+4x
1/2
x
2
solution g(x) =
x
2
+4x
1/2
x
2
= 1+4x
−3/2
. Thus,
dg
dx
= −6x
−5/2
.
June 8, 2011 LTSV SSM Second Pass
104 C HA P T E R 3 DIFFERENTIATION
s(t ) =
1−2t
t
1/2
In Exercises 37–42, calculate the derivative indicated.
37.
dT
dC
¸
¸
¸
C=8
, T = 3C
2/3
solution WithT (C) = 3C
2/3
, wehave
dT
dC
= 2C
−1/3
. Therefore,
dT
dC
¸
¸
¸
¸
C=8
= 2(8)
−1/3
= 1.
dP
dV
¸
¸
¸
V=−2
, P =
7
V
39.
ds
dz
¸
¸
¸
z=2
, s = 4z −16z
2
solution Withs = 4z −16z
2
, wehave
ds
dz
= 4−32z. Therefore,
ds
dz
¸
¸
¸
¸
z=2
= 4−32(2) = −60.
dR
dW
¸
¸
¸
¸
W=1
, R = W
π
41.
dr
dt
¸
¸
¸
¸
t =4
, r =
t
2
+1
t
1/2
solution Wehave
dr
dt
=
d
dt
t
2
+1
t
1/2
=
d
dt
(t
3/2
+t
−1/2
) =
3
2
t
1/2

1
2
t
−3/2
Evaluatingat t = 4gives
dr
dt
¸
¸
¸
¸
t =4
=
3
2
4
1/2

1
2
4
−3/2
=
47
16
dp
dh
¸
¸
¸
¸
h=32
, p = 16h
0.2
+8h
−0.8
43. Matchthefunctions ingraphs (A)–(D) withtheir derivatives (I)–(III) inFigure11. Notethat two of thefunctions
havethesamederivative. Explainwhy.
y
x
x
(A)
y
(I)
x
y
(II)
x
y
(III)
y
x
(B)
y
x
(C)
y
x
(D)
FIGURE 11
solution
• Consider thegraphin(A). Ontheleft sideof thegraph, theslopeof thetangent lineis positivebut ontheright
sidetheslopeof thetangent lineis negative. Thus thederivativeshouldtransitionfrompositivetonegativewith
increasingx. Thismatchesthegraphin(III).
• Consider thegraphin(B). Thisisalinear function, soitsslopeisconstant. Thusthederivativeisconstant, which
matchesthegraphin(I).
• Considerthegraphin(C). Movingfromlefttoright, theslopeof thetangentlinetransitionsfrompositivetonegative
thenback to positive. Thederivativeshouldthereforebenegativeinthemiddleandpositiveto either side. This
matchesthegraphin(II).
• Consider thegraphin(D). Ontheleft sideof thegraph, theslopeof thetangent lineis positivebut ontheright
sidetheslopeof thetangent lineis negative. Thus thederivativeshouldtransitionfrompositivetonegativewith
increasingx. Thismatchesthegraphin(III).
Notethat thefunctionswhosegraphsareshownin(A) and(D) havethesamederivative. Thishappensbecausethe
graphin(D) is just avertical translationof thegraphin(A), whichmeans thetwo functions differ by aconstant. The
derivativeof aconstant iszero, sothetwofunctionsendupwiththesamederivative.
June 8, 2011 LTSV SSM Second Pass
S E C T I ON 3.2 The Derivative as a Function 105
Of thetwofunctionsf andg inFigure12, whichisthederivativeof theother? J ustifyyour answer.
45. Assignthelabelsf (x), g(x), andh(x) tothegraphsinFigure13insuchawaythatf

(x) = g(x) andg

(x) = h(x).
y
x
y
x
y
x
(A) (B) (C)
FIGURE 13
solution Consider thegraphin(A). Movingfromleft toright, theslopeof thetangent lineispositiveover theﬁrst
quarter of thegraph, negativeinthemiddlehalf andpositiveagainover theﬁnal quarter. Thederivativeof thisfunction
must thereforebenegativeinthemiddleandpositiveoneither side. Thismatchesthegraphin(C).
Nowfocus onthegraphin(C). Theslopeof thetangent lineis negativeover theleft half andpositiveontheright
half. Thederivativeof thisfunctionthereforeneedstobenegativeontheleft andpositiveontheright. Thisdescription
matchesthegraphin(B).
Weshould thereforelabel thegraph in (A) as f (x), thegraph in (B) as h(x), and thegraph in (C) as g(x). Then
f

(x) = g(x) andg

(x) = h(x).
Accordingtothepeak oil theory, ﬁrstproposedin1956bygeophysicistM. Hubbert, thetotal amountof crudeoil
Q(t ) producedworldwideuptotimet hasagraphlikethat inFigure14.
(a) SketchthederivativeQ

(t ) for 1900≤ t ≤ 2150. What doesQ

(t ) represent?
(b) Inwhichyear (approximately) doesQ

(t ) takeonitsmaximumvalue?
(c) What isL = lim
t →∞
Q(t )?Andwhat isitsinterpretation?
(d) What isthevalueof lim
t →∞
Q

(t )?
47. Usethetableof valuesof f (x) todeterminewhichof (A) or (B) inFigure15isthegraphof f

(x). Explain.
x 0 0.5 1 1.5 2 2.5 3 3.5 4
f (x) 10 55 98 139 177 210 237 257 268
x
y
x
y
(A) (B)
FIGURE 15 Whichisthegraphof f

(x)?
solution Theincrementbetweensuccessivex valuesinthetableisaconstant0.5buttheincrementbetweensuccessive
f (x) values decreases from45to 43to 41to 38andso on. Thus thedifferencequotients decreasewithincreasingx,
suggestingthat f

(x) decreasesasafunctionof x. Becausethegraphin(B) depictsadecreasingfunction, (B) might be
thegraphof thederivativeof f (x).
Let R beavariableandr aconstant. Computethederivatives:
(a)
d
dR
R (b)
d
dR
r (c)
d
dR
r
2
R
3
49. Computethederivatives, wherec isaconstant.
(a)
d
dt
ct
3
(b)
d
dy
(9c
2
y
3
−24c) (c)
d
dz
(5z +4cz
2
)
solution
(a)
d
dt
ct
3
= 3ct
2
. (b)
d
dz
(5z +4cz
2
) = 5+8cz. (c)
d
dy
(9c
2
y
3
−24c) = 27c
2
y
2
.
Findthepointsonthegraphof f (x) = 12x −x
3
wherethetangent lineishorizontal.
51. Findthepointsonthegraphof y = x
2
+3x −7at whichtheslopeof thetangent lineisequal to4.
solution Let y = x
2
+3x −7. Solvingdy/dx = 2x +3= 4yieldsx =
1
2
.
Findthevaluesof x wherey = x
3
andy = x
2
+5x haveparallel tangent lines.
53. Determinea andb suchthat p(x) = x
2
+ax +b satisﬁesp(1) = 0andp

(1) = 4.
solution Let p(x) = x
2
+ ax + b satisfy p(1) = 0andp

(1) = 4. Now, p

(x) = 2x + a. Therefore0 = p(1) =
1+a +b and4= p

(1) = 2+a; i.e., a = 2andb = −3.
Findall valuesof x suchthat thetangent linetoy = 4x
2
+11x +2issteeper thanthetangent linetoy = x
3
.
55. Letf (x) = x
3
−3x +1. Showthatf

(x) ≥ −3for all x andthat, for everym > −3, therearepreciselytwopoints
wheref

(x) = m. Indicatethepositionof thesepointsandthecorrespondingtangentlinesfor onevalueof minasketch
of thegraphof f (x).
solution Let P = (a, b) beapoint onthegraphof f (x) = x
3
−3x +1.
• Thederivativesatisﬁesf

(x) = 3x
2
−3≥ −3since3x
2
isnonnegative.
• Supposetheslopem of thetangent lineisgreater than−3. Thenf

(a) = 3a
2
−3= m, whence
a
2
=
m+3
3
> 0 andthus a = ±
_
m+3
3
.
• Thetwoparallel tangent lineswithslope2areshownwiththegraphof f (x) here.
June 8, 2011 LTSV SSM Second Pass
106 C HA P T E R 3 DIFFERENTIATION
−2
−1
−2
2
4
1
2
x
y
Showthat thetangent linestoy =
1
3
x
3
−x
2
at x = a andat x = b areparallel if a = b or a +b = 2.
57. Computethederivativeof f (x) = x
3/2
usingthelimit deﬁnition. Hint: Showthat
f (x +h) −f (x)
h
=
(x +h)
3
−x
3
h
_
1
_
(x +h)
3
+

x
3
_
solution Oncewehavethedifferenceof squareroots, wemultiplybytheconjugatetosolvetheproblem.
f

(x) = lim
h→0
(x +h)
3/2
−x
3/2
h
= lim
h→0
_
(x +h)
3

x
3
h
__
(x +h)
3
+

x
3
_
(x +h)
3
+

x
3
_
= lim
h→0
(x +h)
3
−x
3
h
_
1
_
(x +h)
3
+

x
3
_
.
Theﬁrst factor of theexpressioninthelast lineisclearly thelimit deﬁnitionof thederivativeof x
3
, whichis3x
2
. The
secondfactor canbeevaluated, so
d
dx
x
3/2
= 3x
2
1
2

x
3
=
3
2
x
1/2
.
Sketchthegraphof acontinuousfunctionon(0, 5) that isdifferentiableexcept at x = 1andx = 4.
59. Show, usingthelimit deﬁnitionof thederivative, that f (x) = |x
2
−4| isnot differentiableat x = 2.
solution Wehave
f

(2) = lim
h→0
f (2+h) −f (2)
h
= lim
h→0
¸
¸
¸(2+h)
2
−4
¸
¸
¸ −0
h
= lim
h→0
¸
¸
¸h
2
+4h
¸
¸
¸
h
= lim
h→0
|h +4| ·
|h|
h
Ash approacheszero, h +4approaches4. However, thesecondfactor is1if h ispositiveandis−1if h isnegative. Thus
lim
h→0−
|h +4| ·
|h|
h
= −4, and lim
h→0+
|h +4| ·
|h|
h
= 4
Sincethetwoone-sidedlimitsareunequal, f (x) isnot differentiableat x = 2.
Theaveragespeed(inmetersper second) of agasmoleculeis
v
avg
=
_
8RT
πM
whereT is thetemperature(in kelvins), M is themolar mass (in kilograms per mole), and R = 8.31. Calculate
dv
avg
/dT at T = 300K for oxygen, whichhasamolar massof 0.032kg/mol.
61. BiologistshaveobservedthatthepulserateP (inbeatsper minute) inanimalsisrelatedtobodymass(inkilograms)
bytheapproximateformulaP = 200m
−1/4
.Thisisoneof manyallometric scaling laws prevalentinbiology. Is|dP/dm|
anincreasingor decreasingfunctionof m? Findanequationof thetangent lineat thepoints onthegraphinFigure16
that represent goat (m = 33) andman(m = 68).
Mass (kg)
500 400 300 200 100
Cattle
Pulse
(beats/min)
200
100
Guineapig
Goat
Man
FIGURE 16
solution dP/dm = −50m
−5/4
. For m > 0, |dP/dm| = |50m
−5/4
|. |dP/dm| → 0asm getslarger; |dP/dm| gets
smaller asm getsbigger.
For eachm = c, theequationof thetangent linetothegraphof P at m is
y = P

(c)(m−c) +P(c).
For agoat (m = 33kg), P(33) = 83.445beatsper minute(bpm) and
dP
dm
= −50(33)
−5/4
≈ −0.63216bpm/kg.
Hence, y = −0.63216(m−33) +83.445.
June 8, 2011 LTSV SSM Second Pass
S E C T I ON 3.2 The Derivative as a Function 107
For aman(m = 68kg), wehaveP(68) = 69.647bpmand
dP
dm
= −50(68)
−5/4
≈ −0.25606bpm/kg.
Hence, thetangent linehasformulay = −0.25606(m−68) +69.647.
SomestudiessuggestthatkidneymassK inmammals(inkilograms) isrelatedtobodymassm(inkilograms) by
theapproximateformulaK = 0.007m
0.85
. CalculatedK/dm at m = 68. Thencalculatethederivativewithrespect
tom of therelativekidney-to-massratioK/m at m = 68.
63. TheClausius–ClapeyronLawrelatesthevapor pressure of waterP (inatmospheres)tothetemperatureT (inkelvins):
dP
dT
= k
P
T
2
wherek isaconstant. EstimatedP/dT for T = 303, 313, 323, 333, 343usingthedataandtheapproximation
dP
dT

P(T +10) −P(T −10)
20
T (K) 293 303 313 323 333 343 353
P (atm) 0.0278 0.0482 0.0808 0.1311 0.2067 0.3173 0.4754
Doyour estimatesseemtoconﬁrmtheClausius–ClapeyronLaw?What istheapproximatevalueof k?
solution Usingtheindicatedapproximationtotheﬁrst derivative, wecalculate
P

(303) ≈
P(313) −P(293)
20
=
0.0808−0.0278
20
= 0.00265atm/K;
P

(313) ≈
P(323) −P(303)
20
=
0.1311−0.0482
20
= 0.004145atm/K;
P

(323) ≈
P(333) −P(313)
20
=
0.2067−0.0808
20
= 0.006295atm/K;
P

(333) ≈
P(343) −P(323)
20
=
0.3173−0.1311
20
= 0.00931atm/K;
P

(343) ≈
P(353) −P(333)
20
=
0.4754−0.2067
20
= 0.013435atm/K
If theClausius–Clapeyronlawis valid, then
T
2
P
dP
dT
shouldremainconstant as T varies. Usingthedatafor vapor
pressureandtemperatureandtheapproximatederivativevaluescalculatedabove, weﬁnd
T (K) 303 313 323 333 343
T
2
P
dP
dT
5047.59 5025.76 5009.54 4994.57 4981.45
Thesevaluesareroughlyconstant, suggestingthat theClausius–Clapeyronlawisvalid, andthat k ≈ 5000.
Let L bethetangent lineto thehyperbolaxy = 1at x = a, wherea > 0. Showthat theareaof thetriangle
boundedbyL andthecoordinateaxesdoesnot dependona.
65. Inthesettingof Exercise64, showthat thepoint of tangency is themidpoint of thesegment of L lyingintheﬁrst
quadrant.
solution In theprevious exercise, wesawthat thetangent lineto thehyperbolaxy = 1 or y =
1
x
at x = a has
y-intercept P = (0,
2
a
) andx-intercept Q = (2a, 0). Themidpoint of thelinesegment connectingP andQisthus
_
0+2a
2
,
2
a
+0
2
_
=
_
a,
1
a
_
,
whichisthepoint of tangency.
Matchfunctions(A)–(C) withtheir derivatives(I)–(III) inFigure17.
67. Makearoughsketchof thegraphof thederivativeof thefunctioninFigure18(A).
(A) (B)
y = x
2
4 3
4 3 2 1
2 0 1 −1
3
2
1
2
x
y
x
y
FIGURE 18
June 8, 2011 LTSV SSM Second Pass
108 C HA P T E R 3 DIFFERENTIATION
solution Thegraphhas atangent linewithnegativeslopeapproximately ontheinterval (1, 3.6), andhas atangent
linewithapositiveslopeelsewhere. Thisimpliesthatthederivativemustbenegativeontheinterval (1, 3.6) andpositive
elsewhere. Thegraphmaythereforelooklikethis:
y
x
1 2 3 4
Graphthederivativeof thefunctioninFigure18(B), omittingpointswherethederivativeisnot deﬁned.
69. Sketchthegraphof f (x) = x |x|. Thenshowthat f

(0) exists.
solution For x < 0, f (x) = −x
2
, andf

(x) = −2x. For x > 0, f (x) = x
2
, andf

(x) = 2x. At x = 0, weﬁnd
lim
h→0+
f (0+h) −f (0)
h
= lim
h→0+
h
2
h
= 0
and
lim
h→0−
f (0+h) −f (0)
h
= lim
h→0−
−h
2
h
= 0.
Becausethetwoone-sidedlimitsexist andareequal, it followsthat f

(0) existsandisequal tozero. Hereisthegraph
of f (x) = x|x|.
y
x
1 2 −1 −2
2
4
−4
−2
Determinethevaluesof x at whichthefunctioninFigure19is: (a) discontinuous, and(b) nondifferentiable.
In Exercises 71–76, ﬁnd the points c (if any) such that f

(c) does not exist.
71. f (x) = |x −1|
solution
y
x
1
0.5
1
1.5
2
2 3 −1
Hereisthegraphof f (x) = |x −1|. Itsderivativedoesnot exist at x = 1. At that valueof x thereisasharpcorner.
f (x) = [x]
73. f (x) = x
2/3
solution Hereisthegraphof f (x) = x
2/3
. Itsderivativedoesnot exist at x = 0. At that valueof x, thereisasharp
corner or “cusp”.
y
x
1
1
1.5
2 −1 −2
f (x) = x
3/2
75. f (x) = |x
2
−1|
solution Hereisthegraphof f (x) =
¸
¸
¸x
2
−1
¸
¸
¸. Itsderivativedoesnot exist at x = −1or at x = 1. At thesevalues
of x, thegraphhassharpcorners.
June 8, 2011 LTSV SSM Second Pass
S E C T I ON 3.2 The Derivative as a Function 109
y
x
1
2
1
3
2 −1 −2
f (x) = |x −1|
2
In Exercises 77–82, zoom in on a plot of f (x) at the point (a, f (a)) and state whether or not f (x) appears to be
differentiable at x = a. If it is nondifferentiable, state whether the tangent line appears to be vertical or does not exist.
77. f (x) = (x −1)|x|, a = 0
solution Thegraphof f (x) = (x −1)|x| for x near 0isshownbelow. Becausethegraphhasasharpcorner atx = 0,
it appearsthat f isnot differentiableat x = 0. Moreover, thetangent linedoesnot exist at thispoint.
y
x
0.1 0.2 −0.1 −0.2
−0.1
−0.2
−0.3
f (x) = (x −3)
5/3
, a = 3
79. f (x) = (x −3)
1/3
, a = 3
solution Thegraph of f (x) = (x − 3)
1/3
for x near 3 is shown below. Fromthis graph, it appears that f is not
differentiableat x = 3. Moreover, thetangent lineappearstobevertical.
3.05 3.1 2.95 3 2.9
f (x) = sin(x
1/3
), a = 0
81. f (x) = | sinx|, a = 0
solution Thegraphof f (x) = | sinx| for x near 0isshownbelow. Becausethegraphhasasharpcorner at x = 0, it
appearsthat f isnot differentiableat x = 0. Moreover, thetangent linedoesnot exist at thispoint.
y
x
0.05
0.08
0.04
0.1
0.1 −0.05 −0.1
f (x) = |x −sinx|, a = 0
83. Plotthederivativef

(x) of f (x) = 2x
3
−10x
−1
forx > 0(settheboundsof theviewingboxappropriately)
andobservethatf

(x) > 0. Whatdoesthepositivityof f

(x) tell usaboutthegraphof f (x) itself?Plotf (x) andconﬁrm
thisconclusion.
solution Let f (x) = 2x
3
−10x
−1
. Thenf

(x) = 6x
2
+10x
−2
. Thegraphof f

(x) isshownintheﬁgurebelowat
theleft andit isclear that f

(x) > 0for all x > 0. Thepositivityof f

(x) tellsusthat thegraphof f (x) isincreasingfor
x > 0. Thisisconﬁrmedintheﬁgurebelowat theright, whichshowsthegraphof f (x).
8 6 4 2
x
y
100
200
300
400
8 6 4 2
x
y
200
−200
400
600
800
June 8, 2011 LTSV SSM Second Pass
110 C HA P T E R 3 DIFFERENTIATION
Findthecoordinatesof thepoint P inFigure20at whichthetangent linepassesthrough(5, 0).
Exercises 85–88 refer to Figure 21. Length QR is called the subtangent at P, and length RT is called the subnormal.
x
y
P = (x, f (x))
T R
y = f (x)
Q
Tangent line
FIGURE 21
85. Calculatethesubtangent of
f (x) = x
2
+3x at x = 2
solution Let f (x) = x
2
+3x. Thenf

(x) = 2x +3, andtheequationof thetangent lineat x = 2is
y = f

(2)(x −2) +f (2) = 7(x −2) +10= 7x −4.
Thislineintersectsthex-axisat x =
4
7
. ThusQhascoordinates(
4
7
, 0), R hascoordinates(2, 0) andthesubtangent is
2−
4
7
=
10
7
.
Calculatethesubnormal of f (x) = x
2/3
at x = 8.
87. Proveingeneral that thesubnormal at P is|f

(x)f (x)|.
solution Theslopeof thetangent lineat P isf

(x). Theslopeof thelinenormal tothegraphat P isthen−1/f

(x),
andthenormal lineintersectsthex-axisatthepointT withcoordinates(x +f (x)f

(x), 0). ThepointR hascoordinates
(x, 0), sothesubnormal is
|x +f (x)f

(x) −x| = |f (x)f

(x)|.
Showthat PQhaslength|f (x)|
_
1+f

(x)
−2
.
89. ProvethefollowingtheoremofApolloniusof Perga(theGreekmathematicianbornin262bce whogavetheparabola,
ellipse, andhyperbolatheir names): Thesubtangent of theparabolay = x
2
at x = a isequal toa/2.
solution Let f (x) = x
2
. Thetangent linetof at x = a is
y = f

(a)(x −a) +f (a) = 2a(x −a) +a
2
= 2ax −a
2
.
Thex-intercept of thisline(wherey = 0) is
a
2
asclaimed.
y
y = x
2
(a, a
2
)
x
(
–, 0
)
a
2
Showthat thesubtangent toy = x
3
at x = a isequal to
1
3
a.
91. Formulateandproveageneralizationof Exercise90for y = x
n
.
solution Let f (x) = x
n
. Thenf

(x) = nx
n−1
, andtheequationof thetangent linet x = a is
y = f

(a)(x −a) +f (a) = na
n−1
(x −a) +a
n
= na
n−1
x −(n −1)a
n
.
Thislineintersectsthex-axisat x = (n −1)a/n. Thus, Qhascoordinates((n −1)a/n, 0), R hascoordinates(a, 0) and
thesubtangent is
a −
n −1
n
a =
1
n
a.
June 8, 2011 LTSV SSM Second Pass
S E C T I ON 3.2 The Derivative as a Function 111
Further Insights and Challenges
Twosmall archeshavetheshapeof parabolas. Theﬁrstisgivenbyf (x) = 1−x
2
for−1≤ x ≤ 1andthesecond
byg(x) = 4−(x −4)
2
for 2≤ x ≤ 6. A boardisplacedontopof thesearchessoit restsonboth(Figure22). What
istheslopeof theboard? Hint: Findthetangent linetoy = f (x) that intersectsy = g(x) inexactlyonepoint.
FIGURE 22
93. A vaseisformedbyrotatingy = x
2
aroundthey-axis. If wedropinamarble, it will either touchthebottompoint
of thevaseor besuspendedabovethebottombytouchingthesides(Figure23). Howsmall must themarblebetotouch
thebottom?
FIGURE 23
solution Supposeacircleistangenttotheparabolay = x
2
atthepoint(t, t
2
). Theslopeof theparabolaatthispoint
is2t , sotheslopeof theradiusof thecircleat thispoint is−
1
2t
(sinceit isperpendicular tothetangent lineof thecircle).
Thusthecenter of thecirclemust bewherethelinegivenby y = −
1
2t
(x − t ) + t
2
crossesthey-axis. Wecanﬁndthe
y-coordinatebysettingx = 0: weget y =
1
2
+t
2
. Thus, theradiusextendsfrom(0,
1
2
+t
2
) to(t, t
2
) and
r =
_
_
1
2
+t
2
−t
2
_
2
+t
2
=
_
1
4
+t
2
.
Thisradiusisgreater than
1
2
whenever t > 0; so, if amarblehasradius> 1/2it sitsontheedgeof thevase, but if it has
radius≤ 1/2it rollsall thewaytothebottom.
Letf (x) beadifferentiablefunction, andsetg(x) = f (x +c), wherec isaconstant. Usethelimitdeﬁnitionto
showthat g

(x) = f

(x +c). Explainthisresult graphically, recallingthat thegraphof g(x) isobtainedbyshifting
thegraphof f (x) c unitstotheleft (if c > 0) or right (if c < 0).
95. NegativeExponents Letn beawholenumber. UsethePowerRuleforx
n
tocalculatethederivativeof f (x) = x
−n
byshowingthat
f (x +h) −f (x)
h
=
−1
x
n
(x +h)
n
(x +h)
n
−x
n
h
solution Let f (x) = x
−n
wheren isapositiveinteger.
• Thedifferencequotient for f is
f (x +h) −f (x)
h
=
(x +h)
−n
−x
−n
h
=
1
(x+h)
n

1
x
n
h
=
x
n
−(x+h)
n
x
n
(x+h)
n
h
=
−1
x
n
(x +h)
n
(x +h)
n
−x
n
h
.
• Therefore,
f

(x) = lim
h→0
f (x +h) −f (x)
h
= lim
h→0
−1
x
n
(x +h)
n
(x +h)
n
−x
n
h
= lim
h→0
−1
x
n
(x +h)
n
lim
h→0
(x +h)
n
−x
n
h
= −x
−2n
d
dx
_
x
n
_
.
• Fromabove, wecontinue: f

(x) = −x
−2n
d
dx
_
x
n
_
= −x
−2n
· nx
n−1
= −nx
−n−1
. Sincen isapositiveinteger,
k = −n is anegativeinteger andwehave
d
dx
_
x
k
_
=
d
dx
_
x
−n
_
= −nx
−n−1
= kx
k−1
; i.e.
d
dx
_
x
k
_
= kx
k−1
for negativeintegersk.
VerifythePower Rulefor theexponent 1/n, wheren isapositiveinteger, usingthefollowingtrick: Rewritethe
differencequotient for y = x
1/n
at x = b intermsof u = (b +h)
1/n
anda = b
1/n
.
97. InﬁnitelyRapidOscillations Deﬁne
f (x) =

x sin
1
x
x = 0
0 x = 0
Showthat f (x) iscontinuousat x = 0but f

(0) doesnot exist (seeFigure22).
June 8, 2011 LTSV SSM Second Pass
112 C HA P T E R 3 DIFFERENTIATION
solution Let f (x) =
_
x sin
_
1
x
_
if x = 0
0 if x = 0
. Asx → 0,
|f (x) −f (0)| =
¸
¸
¸
¸
x sin
_
1
x
_
−0
¸
¸
¸
¸
= |x|
¸
¸
¸
¸
sin
_
1
x

¸
¸
¸
→ 0
sincethevalues of thesineliebetween −1 and 1. Hence, by theSqueezeTheorem, lim
x→0
f (x) = f (0) and thus f is
continuousat x = 0.
Asx → 0, thedifferencequotient at x = 0,
f (x) −f (0)
x −0
=
x sin
_
1
x
_
−0
x −0
= sin
_
1
x
_
doesnot convergetoalimit sinceit oscillatesinﬁnitelythrougheveryvaluebetween−1and1. Accordingly, f

(0) does
not exist.
For whichvaluesof c doestheequationx
2
+4= cx haveauniquesolution? Hint: Drawagraph.
3.3 Product and Quotient Rules
Preliminary Questions
1. Arethefollowingstatementstrueor false? If false, statethecorrect version.
(a) fg denotesthefunctionwhosevalueat x isf (g(x)).
(b) f/g denotesthefunctionwhosevalueat x isf (x)/g(x).
(c) Thederivativeof theproduct istheproduct of thederivatives.
(d)
d
dx
(fg)
¸
¸
¸
¸
x=4
= f (4)g

(4) −g(4)f

(4)
(e)
d
dx
(fg)
¸
¸
¸
¸
x=0
= f (0)g

(0) +g(0)f

(0)
solution
(a) False. Thenotationfg denotesthefunctionwhosevalueat x isf (x)g(x).
(b) True.
(c) False. Thederivativeof aproduct fg isf

(x)g(x) +f (x)g

(x).
(d) False.
d
dx
(fg)
¸
¸
¸
¸
x=4
= f (4)g

(4) +g(4)f

(4).
(e) True.
2. Find(f/g)

(1) if f (1) = f

(1) = g(1) = 2andg

(1) = 4.
solution
d
dx
(f/g)
¸
¸
x=1
= [g(1)f

(1) −f (1)g

(1)]/g(1)
2
= [2(2) −2(4)]/2
2
= −1.
3. Findg(1) if f (1) = 0, f

(1) = 2, and(fg)

(1) = 10.
solution (fg)

(1) = f (1)g

(1) +f

(1)g(1), so10= 0· g

(1) +2g(1) andg(1) = 5.
Exercises
In Exercises 1–6, use the Product Rule to calculate the derivative.
1. f (x) = x
3
(2x
2
+1)
solution Let f (x) = x
3
(2x
2
+1). Then
f

(x) = x
3
d
dx
(2x
2
+1) +(2x
2
+1)
d
dx
x
3
= x
3
(4x) +(2x
2
+1)(3x
2
) = 10x
4
+3x
2
.
f (x) = (3x −5)(2x
2
−3)
3. f (x) =

x(1−x
3
)
solution Let f (x) =

x(1−x
3
). Then
f

(x) =

x
d
dx
(1−x
3
) +(1−x
3
)
d
dx

x =

x(−3x
2
) +(1−x
3
)
_
1
2
x
−1/2
_
= −3x
5/2
+
1
2
x
−1/2

1
2
x
5/2
= −
7
2
x
5/2
+
1
2x
1/2
June 8, 2011 LTSV SSM Second Pass
S E C T I ON 3.3 Product and Quotient Rules 113
f (x) = (3x
4
+2x
6
)(x −2)
5.
dh
ds
¸
¸
¸
¸
s=4
, h(s) = (s
−1/2
+2s)(7−s
−1
)
solution Let h(s) = (s
−1/2
+2s)(7−s
−1
). Then
dh
ds
= (s
−1/2
+2s)
d
dx
(7−s
−1
) +(7−s
−1
)
d
ds
_
s
−1/2
+2s
_
= (s
−1/2
+2s)(s
−2
) +(7−s
−1
)
_

1
2
s
−3/2
+2
_
= −
7
2
s
−3/2
+
3
2
s
−5/2
+14.
Therefore,
dh
ds
¸
¸
¸
¸
s=4
= −
7
2
(4)
−3/2
+
3
2
(4)
−5/2
+14=
871
64
.
y = (t −8t
−1
)(t +t
2
)
In Exercises 7–12, use the Quotient Rule to calculate the derivative.
7. f (x) =
x
x −2
solution Let f (x) =
x
x−2
. Then
f

(x) =
(x −2)
d
dx
x −x
d
dx
(x −2)
(x −2)
2
=
(x −2) −x
(x −2)
2
=
−2
(x −2)
2
.
f (x) =
x +4
x
2
+x +1
9.
dg
dt
¸
¸
¸
¸
t =−2
, g(t ) =
t
2
+1
t
2
−1
solution Let g(t ) =
t
2
+1
t
2
−1
. Then
dg
dt
=
(t
2
−1)
d
dt
(t
2
+1) −(t
2
+1)
d
dt
(t
2
−1)
(t
2
−1)
2
=
(t
2
−1)(2t ) −(t
2
+1)(2t )
(t
2
−1)
2
= −
4t
(t
2
−1)
2
.
Therefore,
dg
dt
¸
¸
¸
¸
t =−2
= −
4(−2)
((−2)
2
−1)
2
=
8
9
.
dw
dz
¸
¸
¸
¸
z=9
, w =
z
2

z +z
11. g(x) =
1
1+x
3/2
solution
g

(x) =
(1+x
3/2
)
d
dt
(1) −1
d
dt
(1+x
3/2
)
(1+x
3/2
)
2
= −
3
2
x
1/2
(1+x
3/2
)
2
= −
3

x
2(1+x
3/2
)
2
h(s) =
s
3/2
s
2
+1
In Exercises 13–16, calculate the derivative in two ways. First use the Product or Quotient Rule; then rewrite the function
algebraically and apply the Power Rule directly.
13. f (t ) = (2t +1)(t
2
−2)
solution Let f (t ) = (2t +1)(t
2
−2). Then, usingtheProduct Rule,
f

(t ) = (2t +1)(2t ) +(t
2
−2)(2) = 6t
2
+2t −4.
Multiplyingout ﬁrst, weﬁndf (t ) = 2t
3
+t
2
−4t −2. Therefore, f

(t ) = 6t
2
+2t −4.
f (x) = x
2
(3+x
−1
) 15. h(t ) =
t
2
−1
t −1
solution Let h(t ) =
t
2
−1
t −1
. Usingthequotient rule,
f

(t ) =
(t −1)(2t ) −(t
2
−1)(1)
(t −1)
2
=
t
2
−2t +1
(t −1)
2
= 1
for t = 1. Simplifyingﬁrst, weﬁndfor t = 1,
h(t ) =
(t −1)(t +1)
(t −1)
= t +1.
Henceh

(t ) = 1for t = 1.
June 8, 2011 LTSV SSM Second Pass
114 C HA P T E R 3 DIFFERENTIATION
g(x) =
x
3
+2x
2
+3x
−1
x
In Exercises 17–38, calculate the derivative.
17. f (x) = (x
3
+5)(x
3
+x +1)
solution Let f (x) = (x
3
+5)(x
3
+x +1). Then
f

(x) = (x
3
+5)(3x
2
+1) +(x
3
+x +1)(3x
2
) = 6x
5
+4x
3
+18x
2
+5.
f (x) =
_
1
x
−x
2
_
(x
3
+1)
19.
dy
dx
¸
¸
¸
¸
x=3
, y =
1
x +10
solution Let y =
1
x+10
. Usingthequotient rule:
dy
dx
=
(x +10)(0) −1(1)
(x +10)
2
= −
1
(x +10)
2
.
Therefore,
dy
dx
¸
¸
¸
¸
x=3
= −
1
(3+10)
2
= −
1
169
.
dz
dx
¸
¸
¸
¸
x=−2
, z =
x
3x
2
+1
21. f (x) = (

x +1)(

x −1)
solution Let f (x) = (

x + 1)(

x − 1). Multiplying through ﬁrst yields f (x) = x − 1 for x ≥ 0. Therefore,
f

(x) = 1for x ≥ 0. If wecarryout theproduct ruleonf (x) = (x
1/2
+1)(x
1/2
−1), weget
f

(x) = (x
1/2
+1)
_
1
2
(x
−1/2
)
_
+(x
1/2
−1)
_
1
2
x
−1/2
_
=
1
2
+
1
2
x
−1/2
+
1
2

1
2
x
−1/2
= 1.
f (x) =
9x
5/2
−2
x
23.
dy
dx
¸
¸
¸
¸
x=2
, y =
x
4
−4
x
2
−5
solution Let y =
x
4
−4
x
2
−5
. Then
dy
dx
=
_
x
2
−5
_ _
4x
3
_

_
x
4
−4
_
(2x)
_
x
2
−5
_
2
=
2x
5
−20x
3
+8x
_
x
2
−5
_
2
.
Therefore,
dy
dx
¸
¸
¸
¸
x=2
=
2(2)
5
−20(2)
3
+8(2)
(2
2
−5)
2
= −80.
f (x) =
x
4
+x
−1
x +1
25.
dz
dx
¸
¸
¸
¸
x=1
, z =
1
x
3
+1
solution Let z =
1
x
3
+1
. Usingthequotient rule:
dz
dx
=
(x
3
+1)(0) −1(3x
2
)
(x
3
+1)
2
= −
3x
2
(x
3
+1)
2
.
Therefore,
dz
dx
¸
¸
¸
¸
x=1
= −
3(1)
2
(1
3
+1)
2
= −
3
4
.
f (x) =
3x
3
−x
2
+2

x
27. h(t ) =
t
(t +1)(t
2
+1)
solution Let h(t ) =
t
(t +1)(t
2
+1)
=
t
t
3
+t
2
+t +1
. Then
h

(t ) =
_
t
3
+t
2
+t +1
_
(1) −t
_
3t
2
+2t +1
_
_
t
3
+t
2
+t +1
_
2
=
−2t
3
−t
2
+1
_
t
3
+t
2
+t +1
_
2
.
f (x) = x
3/2
_
2x
4
−3x +x
−1/2
_ 29. f (t ) = 3
1/2
· 5
1/2
solution Let f (t ) =

3

5. Thenf

(t ) = 0, sincef (t ) isaconstant function!
June 8, 2011 LTSV SSM Second Pass
S E C T I ON 3.3 Product and Quotient Rules 115
h(x) = π
2
(x −1)
31. f (x) = (x +3)(x −1)(x −5)
solution Let f (x) = (x + 3)(x − 1)(x − 5). UsingtheProduct RuleinsidetheProduct Rulewithaﬁrst factor of
(x +3) andasecondfactor of (x −1)(x −5), weﬁnd
f

(x) = (x +3) ((x −1)(1) +(x −5)(1)) +(x −1)(x −5)(1) = 3x
2
−6x −13.
Alternatively,
f (x) = (x +3)
_
x
2
−6x +5
_
= x
3
−3x
2
−13x +15.
Therefore, f

(x) = 3x
2
−6x −13.
h(s) = s(s +4)(s
2
+1) 33. f (x) =
x
3/2
(x
2
+1)
x +1
solution Usingthequotient rule, andthenusingtheproduct ruletodifferentiatethenumerator, weﬁnd
f

(x) =
(x +1)
_
3
2
x
1/2
(x
2
+1) +x
3/2
(2x)
_
−x
3/2
(x
2
+1)(1)
(x +1)
2
=
5x
7/2
+7x
5/2
+x
3/2
+3x
1/2
2(x +1)
2
g(z) =
(z −2)(z
2
+1)
z
35. g(z) =
_
z
2
−4
z −1
__
z
2
−1
z +2
_
Hint: Simplifyﬁrst.
solution Let
g(z) =
_
z
2
−4
z −1
__
z
2
−1
z +2
_
=
_
(z +2)(z −2)
z −1
__
(z +1)(z −1)
z +2
_
= (z −2)(z +1)
for z = −2andz = 1. Then,
g

(z) = (z +1)(1) +(z −2)(1) = 2z −1.
d
dx
_
(ax +b)(abx
2
+1)
_
(a, b constants)
37.
d
dt
_
xt −4
t
2
−x
_
(x constant)
solution Let f (t ) =
xt −4
t
2
−x
. Usingthequotient rule:
f

(t ) =
(t
2
−x)(x) −(xt −4)(2t )
(t
2
−x)
2
=
xt
2
−x
2
−2xt
2
+8t
(t
2
−x)
2
=
−xt
2
+8t −x
2
(t
2
−x)
2
.
d
dx
_
ax +b
cx +d
_
(a, b, c, d constants)
In Exercises 39–42, calculate the derivative using the values:
f (4) f

(4) g(4) g

(4)
10 −2 5 −1
39. (fg)

(4) and(f/g)

(4).
solution Let h = fg andH = f/g. Thenh

= fg

+gf

andH

=
gf

−fg

g
2
. Finally,
h

(4) = f (4)g

(4) +g(4)f

(4) = (10)(−1) +(5)(−2) = −20,
and
H

(4) =
g(4)f

(4) −f (4)g

(4)
(g(4))
2
=
(5)(−2) −(10)(−1)
(5)
2
= 0.
F

(4), whereF(x) = x
2
f (x).
41. G

(4), whereG(x) = g(x)
2
.
solution Let G(x) = g(x)
2
= g(x)g(x). ThenG

(x) = g(x)g

(x) +g(x)g

(x) = 2g(x)g

(x), and
G

(4) = 2g(4)g

(4) = 2(5)(−1) = −10.
June 8, 2011 LTSV SSM Second Pass
116 C HA P T E R 3 DIFFERENTIATION
H

(4), whereH(x) =
x
g(x)f (x)
.
43. CalculateF

(0), where
F(x) =
x
9
+x
8
+4x
5
−7x
x
4
−3x
2
+2x +1
Hint: DonotcalculateF

(x). Instead, writeF(x) = f (x)/g(x) andexpressF

(0) directlyintermsof f (0), f

(0), g(0),
g

(0).
solution Takingthehint, let
f (x) = x
9
+x
8
+4x
5
−7x
andlet
g(x) = x
4
−3x
2
+2x +1.
ThenF(x) =
f (x)
g(x)
. Now,
f

(x) = 9x
8
+8x
7
+20x
4
−7 and g

(x) = 4x
3
−6x +2.
Moreover, f (0) = 0, f

(0) = −7, g(0) = 1, andg

(0) = 2.
Usingthequotient rule:
F

(0) =
g(0)f

(0) −f (0)g

(0)
(g(0))
2
=
−7−0
1
= −7.
ProceedasinExercise43tocalculateF

(0), where
F(x) =
_
1+x +x
4/3
+x
5/3
_ 3x
5
+5x
4
+5x +1
8x
9
−7x
4
+1
45. Verifytheformula(x
3
)

= 3x
2
bywritingx
3
= x · x · x andapplyingtheProduct Rule.
solution Usingtheproduct rule, wehave
(x
3
)

= ((x · x) · x)

= (x · x) · x

+(x · x)

· x = (x · x) +(x · x

+x

· x) · x = (x · x) +(2x) · x = 3x
2
Plot the derivative of f (x) = x/(x
2
+1) over [−4, 4]. Use the graph to determine the intervals on which
f

(x) > 0andf

(x) < 0. Thenplot f (x) anddescribehowthesignof f

(x) isreﬂectedinthegraphof f (x).
47. Plot f (x) = x/(x
2
−1) (in asuitably bounded viewing box). Usetheplot to determinewhether f

(x) is
positiveor negativeonitsdomain{x : x = ±1}. Thencomputef

(x) andconﬁrmyour conclusionalgebraically.
solution Let f (x) =
x
x
2
−1
. Thegraphof f (x) isshownbelow. Fromthisplot, weseethat f (x) isdecreasingon
itsdomain{x : x = ±1}. Consequently, f

(x) must benegative. Usingthequotient rule, weﬁnd
f

(x) =
(x
2
−1)(1) −x(2x)
(x
2
−1)
2
= −
x
2
+1
(x
2
−1)
2
,
whichisnegativefor all x = ±1.
4 3 2 1
x
y
5
−5
−1 −2 −3 −4
Let P = V
2
R/(R +r)
2
asinExample7. CalculatedP/dr, assumingthat r isvariableandR isconstant.
49. Findall valuesof a suchthat thetangent lineto
f (x) =
x −1
x +8
at x = a
passesthroughtheorigin(Figure4).
−8 −4 4 8
y
x
FIGURE 4
June 8, 2011 LTSV SSM Second Pass
S E C T I ON 3.3 Product and Quotient Rules 117
solution Theslopeof thetangent lineisgivenbyf

(x), whichis
f

(x) =
(x +8)(1) −(x −1)(1)
(x +8)
2
=
9
(x +8)
2
Whenx = a, then, wehavef (a) =
a −1
a +8
andf

(a) =
9
(a+8)
2
, sothat theequationof thetangent lineis
y = f

(a)(x −a) +f (a) =
9
(a +8)
2
(x −a) +
a −1
a +8
=
2
(a +8)
2
x +
−9a +(a −1)(a +8)
(a +8)
2
=
2
(a +8)
2
x +
a
2
−2a −8
(a +8)
2
Thetangent linepassesthroughtheoriginwhenitsconstant termiszero, i.e. whena
2
−2a −8= 0, or for a = 4, −2.
CurrentI (amperes), voltageV (volts), andresistanceR (ohms) inacircuitarerelatedbyOhm’sLaw, I = V/R.
(a) Calculate
dI
dR
¸
¸
¸
¸
R=6
if V isconstant withvalueV = 24.
(b) Calculate
dV
dR
¸
¸
¸
¸
R=6
if I isconstant withvalueI = 4.
51. Therevenueper monthearnedbytheCoutureclothingchainattimet isR(t ) = N(t )S(t ), whereN(t ) isthenumber
of stores andS(t ) is averagerevenueper storeper month. Coutureembarks onatwo-part campaign: (A) to buildnew
storesatarateof 5storesper month, and(B) touseadvertisingtoincreaseaveragerevenueper storeatarateof \$10,000
per month. Assumethat N(0) = 50andS(0) = \$150,000.
(a) Showthat total revenuewill increaseat therate
dR
dt
= 5S(t ) +10,000N(t )
Notethat thetwotermsintheProduct Rulecorrespondtotheseparateeffectsof increasingthenumber of storesonthe
onehand, andtheaveragerevenueper storeontheother.
(b) Calculate
dR
dt
¸
¸
¸
¸
t =0
.
(c) If Couturecanimplement only oneleg(A or B) of its expansionat t = 0, whichchoicewill growrevenuemost
rapidly?
solution
(a) GivenR(t ) = N(t )S(t ), it followsthat
dR
dt
= N(t )S

(t ) +S(t )N

(t ).
Wearetoldthat N

(t ) = 5storesper monthandS

(t ) = 10,000dollarsper month. Therefore,
dR
dt
= 5S(t ) +10,000N(t ).
(b) Usingpart (a) andthegivenvaluesof N(0) andS(0), weﬁnd
dR
dt
¸
¸
¸
¸
t =0
= 5(150,000) +10,000(50) = 1,250,000.
(c) Frompart (b), weseethat of thetwo terms contributingto total revenuegrowth, theterm5S(0) is larger thanthe
term10,000N(0). Thus, if onlyonelegof thecampaigncanbeimplemented, it shouldbepartA: increasethenumber of
storesby5per month.
Thetipspeedratioof aturbine(Figure5) istheratioR = T/W, whereT isthespeedof thetipof abladeand
W isthespeedof thewind. (Engineershavefoundempiricallythataturbinewithn bladesextractsmaximumpower
fromthewindwhenR = 2π/n.) CalculatedR/dt (t inminutes) if W = 35km/handW decreases at arateof 4
km/hper minute, andthetipspeedhasconstant valueT = 150km/h.
53. The curve y = 1/(x
2
+1) is called the witch of Agnesi (Figure 6) after the Italian mathematician MariaAgnesi
(1718–1799), whowroteoneof theﬁrstbooksoncalculus. Thisstrangenameistheresultof amistranslationof theItalian
wordla versiera, meaning“that whichturns.” Findequationsof thetangent linesat x = ±1.
3 2 1 −2 −3 −1
1
x
y
FIGURE 6 Thewitchof Agnesi.
solution Let f (x) =
1
x
2
+1
. Thenf

(x) =
(x
2
+1)(0) −1(2x)
(x
2
+1)
2
= −
2x
_
x
2
+1
_
2
.
• At x = −1, thetangent lineis
y = f

(−1)(x +1) +f (−1) =
1
2
(x +1) +
1
2
=
1
2
x +1.
June 8, 2011 LTSV SSM Second Pass
118 C HA P T E R 3 DIFFERENTIATION
• At x = 1, thetangent lineis
y = f

(1)(x −1) +f (1) = −
1
2
(x −1) +
1
2
= −
1
2
x +1.
Let f (x) = g(x) = x. Showthat (f/g)

= f

/g

.
55. UsetheProduct Ruletoshowthat (f
2
)

= 2ff

.
solution Let g = f
2
= ff . Theng

=
_
f
2
_

= (ff )

= ff

+ff

= 2ff

.
Showthat (f
3
)

= 3f
2
f

.
Further Insights and Challenges
57. Let f , g, h bedifferentiablefunctions. Showthat (fgh)

(x) isequal to
f (x)g(x)h

(x) +f (x)g

(x)h(x) +f

(x)g(x)h(x)
Hint: Writefgh asf (gh).
solution Let p = fgh. Then
p

= (fgh)

= f
_
gh

+hg

_
+ghf

= f

gh +fg

h +fgh

.
ProvetheQuotient Ruleusingthelimit deﬁnitionof thederivative.
59. Derivativeof theReciprocal Usethelimit deﬁnitiontoprove
d
dx
_
1
f (x)
_
= −
f

(x)
f
2
(x)
7
Hint: Showthat thedifferencequotient for 1/f (x) isequal to
f (x) −f (x +h)
hf (x)f (x +h)
solution Let g(x) =
1
f (x)
. Wethencomputethederivativeof g(x) usingthedifferencequotient:
g

(x) = lim
h→0
g(x +h) −g(x)
h
= lim
h→0
1
h
_
1
f (x +h)

1
f (x)
_
= lim
h→0
1
h
_
f (x) −f (x +h)
f (x)f (x +h)
_
= − lim
h→0
_
f (x +h) −f (x)
h
__
1
f (x)f (x +h)
_
.
Wecan apply theruleof products for limits. Theﬁrst parenthetical expression is thedifferencequotient deﬁnition of
f

(x). Thesecondcanbeevaluatedat h = 0togive
1
(f (x))
2
. Hence
g

(x) =
d
dx
_
1
f (x)
_
= −
f

(x)
f
2
(x)
.
ProvetheQuotient RuleusingEq. (7) andtheProduct Rule.
61. Usethelimit deﬁnitionof thederivativetoprovethefollowingspecial caseof theProduct Rule:
d
dx
(xf (x)) = xf

(x) +f (x)
solution First notethat becausef (x) isdifferentiable, it isalsocontinuous. It followsthat
lim
h→0
f (x +h) = f (x).
Nowwetacklethederivative:
d
dx
(xf (x)) = lim
h→0
(x +h)f (x +h) −f (x)
h
= lim
h→0
_
x
f (x +h) −f (x)
h
+f (x +h)
_
= x lim
h→0
f (x +h) −f (x)
h
+ lim
h→0
f (x +h)
= xf

(x) +f (x).
Carryout MariaAgnesi’sproof of theQuotient Rulefromher book oncalculus, publishedin1748: Assumethat
f , g, andh = f/g aredifferentiable. Computethederivativeof hg = f usingtheProduct Rule, andsolvefor h

.
63. ThePower RuleRevisited If youarefamiliar withproof by induction, useinductiontoprovethePower Rulefor
all wholenumbersn. Showthat thePower Ruleholdsfor n = 1; thenwritex
n
asx · x
n−1
andusetheProduct Rule.
solution Let k beapositiveinteger. If k = 1, thenx
k
= x. Notethat
d
dx
_
x
1
_
=
d
dx
(x) = 1= 1x
0
.
June 8, 2011 LTSV SSM Second Pass
S E C T I ON 3.4 Rates of Change 119
HencethePower Ruleholdsfor k = 1. Assumeit holdsfor k = n wheren ≥ 2. Thenfor k = n +1, wehave
d
dx
_
x
k
_
=
d
dx
_
x
n+1
_
=
d
dx
_
x · x
n
_
= x
d
dx
_
x
n
_
+x
n
d
dx
(x)
= x · nx
n−1
+x
n
· 1= (n +1)x
n
= kx
k−1
Accordingly, thePower Ruleholdsfor all positiveintegersbyinduction.
Exercises 64 and 65: A basic fact of algebra states that c is a root of a polynomial f (x) if and only if f (x) = (x −c)g(x)
for some polynomial g(x). We say that c is a multiple root if f (x) = (x −c)
2
h(x), where h(x) is a polynomial.
Showthat c isamultipleroot of f (x) if andonlyif c isaroot of bothf (x) andf

(x).
65. UseExercise64todeterminewhether c = −1isamultipleroot:
(a) x
5
+2x
4
−4x
3
−8x
2
−x +2
(b) x
4
+x
3
−5x
2
−3x +2
solution
(a) Toshowthat −1isamultipleroot of
f (x) = x
5
+2x
4
−4x
3
−8x
2
−x +2,
it sufﬁcestocheckthat f (−1) = f

(−1) = 0. Wehavef (−1) = −1+2+4−8+1+2= 0and
f

(x) = 5x
4
+8x
3
−12x
2
−16x −1
f

(−1) = 5−8−12+16−1= 0
(b) Let f (x) = x
4
+x
3
−5x
2
−3x +2. Thenf

(x) = 4x
3
+3x
2
−10x −3. Because
f (−1) = 1−1−5+3+2= 0
but
f

(−1) = −4+3+10−3= 6= 0,
it followsthat x = −1isaroot of f , but not amultipleroot.
Figure7isthegraphof apolynomial withrootsatA, B, andC. Whichof theseisamultipleroot?Explainyour
reasoningusingExercise64.
3.4 Rates of Change
Preliminary Questions
1. Whichunitsmight beusedfor eachrateof change?
(a) Pressure(inatmospheres) inawater tankwithrespect todepth
(b) Therateof achemical reaction(changeinconcentrationwithrespect totimewithconcentrationinmolesper liter)
solution
(a) Therateof changeof pressurewithrespect todepthmight bemeasuredinatmospheres/meter.
(b) Thereactionrateof achemical reactionmight bemeasuredinmoles/(liter·hour).
2. Twotrainstravel fromNewOrleanstoMemphisin4hours. Theﬁrst traintravelsat aconstant velocityof 90mph,
but thevelocityof thesecondtrainvaries. What wasthesecondtrain’saveragevelocityduringthetrip?
solution Sincebothtrainstravel thesamedistanceinthesameamount of time, theyhavethesameaveragevelocity:
90mph.
3. Estimatef (26), assumingthat f (25) = 43, f

(25) = 0.75.
solution f (x) ≈ f (25) +f

(25)(x −25), sof (26) ≈ 43+0.75(26−25) = 43.75.
4. ThepopulationP(t ) of Freedoniain2009wasP(2009) = 5million.
(a) What isthemeaningof P

(2009)?
(b) EstimateP(2010) if P

(2009) = 0.2.
solution
(a) BecauseP(t ) measuresthepopulationof Freedoniaasafunctionof time, thederivativeP

(2009) measurestherate
of changeof thepopulationof Freedoniaintheyear 2009.
(b) P(2010) ≈ P(2009) +P

(2010). Thus, if P

(2009) = 0.2, thenP(2009) ≈ 5.2million.
June 8, 2011 LTSV SSM Second Pass
120 C HA P T E R 3 DIFFERENTIATION
Exercises
In Exercises 1–8, ﬁnd the rate of change.
1. Areaof asquarewithrespect toitssides whens = 5.
solution Let theareabeA = f (s) = s
2
. Thentherateof changeof A withrespect to s is d/ds(s
2
) = 2s. When
s = 5, theareachangesat arateof 10squareunitsper unit increase. (Drawa5×5squareongraphpaper andtracethe
areaaddedbyincreasingeachsidelengthby1, excludingthecorner, toseewhat thismeans.)
Volumeof acubewithrespect toitssides whens = 5.
3. Cuberoot
3

x withrespect tox whenx = 1, 8, 27.
solution Let f (x) =
3

x. Writingf (x) = x
1/3
, weseetherateof changeof f (x) withrespect to x is givenby
f

(x) =
1
3
x
−2/3
. Therequestedratesof changearegiveninthetablethat follows:
c ROC of f (x) withrespect tox at x = c.
1 f

(1) =
1
3
(1) =
1
3
8 f

(8) =
1
3
(8
−2/3
) =
1
3
(
1
4
) =
1
12
27 f

(27) =
1
3
(27
−2/3
) =
1
3
(
1
9
) =
1
27
Thereciprocal 1/x withrespect tox whenx = 1, 2, 3.
5. Thediameter of acirclewithrespect toradius.
solution Therelationshipbetweenthediameter d of acircleandits radius r is d = 2r. Therateof changeof the
diameter withrespect totheradiusisthend

= 2.
SurfaceareaA of aspherewithrespect toradiusr (A = 4πr
2
).
7. VolumeV of acylinder withrespect toradiusif theheight isequal totheradius.
solution Thevolumeof thecylinder isV = πr
2
h = πr
3
. ThusdV/dr = 3πr
2
.
Speedof soundv (inm/s) withrespect toair temperatureT (inkelvins), wherev = 20

T .
In Exercises 9–11, refer to Figure 10, the graph of distance s(t ) from the origin as a function of time for a car trip.
t (h)
3.0 2.5 2.0 1.5 1.0 0.5
150
100
50
Distance(km)
FIGURE 10 Distancefromtheoriginversustimefor acar trip.
9. Findtheaveragevelocityover eachinterval.
(a) [0, 0.5] (b) [0.5, 1] (c) [1, 1.5] (d) [1, 2]
solution
(a) Theaveragevelocityover theinterval [0, 0.5] is
50−0
0.5−0
= 100km/hour.
(b) Theaveragevelocityover theinterval [0.5, 1] is
100−50
1−0.5
= 100km/hour.
(c) Theaveragevelocityover theinterval [1, 1.5] is
100−100
1.5−1
= 0km/hour.
(d) Theaveragevelocityover theinterval [1, 2] is
50−100
2−1
= −50km/hour.
At what timeisvelocityat amaximum?
11. Matchthedescriptions(i)–(iii) withtheintervals(a)–(c).
(i) Velocityincreasing
(ii) Velocitydecreasing
(iii) Velocitynegative
(a) [0, 0.5]
(b) [2.5, 3]
(c) [1.5, 2]
June 8, 2011 LTSV SSM Second Pass
S E C T I ON 3.4 Rates of Change 121
solution
(a) (i): Thedistancecurveisincreasing, andisalsobending upward, sothat distanceisincreasingat anincreasingrate.
(b) (ii) : Over theinterval [2.5, 3], thedistancecurveis ﬂattening, showing that thecar is slowing down; that is, the
velocityisdecreasing.
(c) (iii): Thedistancecurveisdecreasing, sothetangent linehasnegativeslope; thismeansthevelocityisnegative.
UsethedatafromTable1inExample1tocalculatetheaveragerateof changeof MartiantemperatureT with
respect totimet over theinterval from8:36am to9:34am.
13. UseFigure3fromExample1to estimatetheinstantaneous rateof changeof Martiantemperaturewithrespect to
time(indegreesCelsiusper hour) at t = 4am.
solution Thesegmentof thetemperaturegrapharoundt = 4am appearstobeastraightlinepassingthroughroughly
(1:36, −70) and(4:48, −75). Theinstantaneousrateof changeof Martiantemperaturewithrespect totimeat t = 4am
isthereforeapproximately
dT
dt
=
−75−(−70)
3.2
= −1.5625

C/hour.
Thetemperature(in

C) of anobject at timet (inminutes) isT (t ) =
3
8
t
2
− 15t + 180for 0≤ t ≤ 20. At what
rateistheobject coolingat t = 10? (Givecorrect units.)
15. Thevelocity(incm/s)of bloodmoleculesﬂowingthroughacapillaryof radius0.008cmisv = 6.4×10
−8
−0.001r
2
,
wherer isthedistancefromthemoleculetothecenter of thecapillary. Findtherateof changeof velocity withrespect
tor whenr = 0.004cm.
solution Therateof changeof thevelocityof thebloodmoleculesisv

(r) = −0.002r. Whenr = 0.004cm, thisrate
is−8×10
−6
cm/s.
Figure11displays thevoltageV across acapacitor as afunctionof timewhilethecapacitor is beingcharged.
Estimatetherateof changeof voltageat t = 20s. Indicatethevaluesinyour calculationandincludeproper units.
Doesvoltagechangemorequicklyor moreslowlyastimegoeson? Explainintermsof tangent lines.
17. UseFigure12toestimatedT /dh at h = 30and70, whereT isatmospherictemperature(indegreesCelsius) andh
isaltitude(inkilometers). WhereisdT /dh equal tozero?
T
r
o
p
o
s
p
h
e
r
e
S
t
r
a
t
o
s
p
h
e
r
e
M
e
s
o
s
p
h
e
r
e
T
h
e
r
m
o
s
p
h
e
r
e
Altitude(km)
Temperature
(˚C)
250
200
150
100
50
0
−50
−100
10 50 100 150
FIGURE 12 Atmospherictemperatureversusaltitude.
solution Ath = 30km,thegraphof atmospherictemperatureappearstobelinearpassingthroughthepoints(23, −50)
and(40, 0). Theslopeof thissegment of thegraphisthen
0−(−50)
40−23
=
50
17
= 2.94;
so
dT
dh
¸
¸
¸
¸
h=30
≈ 2.94

C/km.
Ath = 70km,thegraphof atmospherictemperatureappearstobelinearpassingthroughthepoints(58, 0) and(88, −100).
Theslopeof thissegment of thegraphisthen
−100−0
88−58
=
−100
30
= −3.33;
so
dT
dh
¸
¸
¸
¸
h=70
≈ −3.33

C/km.
dT
dh
= 0atthosepointswherethetangentlineonthegraphishorizontal. Thisappearstohappenovertheinterval [13, 23],
andnear thepointsh = 50andh = 90.
Theearthexerts agravitational forceof F(r) = (2.99×10
16
)/r
2
newtons onanobject withamass of 75kg
locatedr metersfromthecenter of theearth. Findtherateof changeof forcewithrespecttodistancer atthesurface
of theearth.
19. Calculatetherateof changeof escapevelocity v
esc
= (2.82× 10
7
)r
−1/2
m/s withrespect to distancer fromthe
center of theearth.
solution Theratethat escapevelocitychangesisv

esc
(r) = −1.41×10
7
r
−3/2
.
Thepower deliveredbyabatterytoanapparatusof resistanceR (inohms) isP = 2.25R/(R +0.5)
2
watts. Find
therateof changeof power withrespect toresistancefor R = 3 andR = 5.
June 8, 2011 LTSV SSM Second Pass
122 C HA P T E R 3 DIFFERENTIATION
21. Thepositionof aparticlemovinginastraight lineduringa5-stripiss(t ) = t
2
−t +10cm. Findatimet at which
theinstantaneousvelocityisequal totheaveragevelocityfor theentiretrip.
solution Let s(t ) = t
2
− t + 10, 0 ≤ t ≤ 5, withs incentimeters (cm) andt inseconds (s). Theaveragevelocity
over thet -interval [0, 5] is
s(5) −s(0)
5−0
=
30−10
5
= 4cm/s.
The (instantaneous) velocity is v(t ) = s

(t ) = 2t − 1. Solving 2t − 1 = 4 yields t =
5
2
s, the time at which the
instantaneousvelocityequalsthecalculatedaveragevelocity.
Theheight (inmeters) of ahelicopter at timet (inminutes) iss(t ) = 600t −3t
3
for 0≤ t ≤ 12.
(a) Plot s(t ) andvelocityv(t ).
(b) Findthevelocityat t = 8andt = 10.
(c) Findthemaximumheight of thehelicopter.
23. A particlemovingalongalinehaspositions(t ) = t
4
− 18t
2
mat timet seconds. At whichtimesdoestheparticle
passthroughtheorigin?At whichtimesistheparticleinstantaneouslymotionless(that is, it haszerovelocity)?
solution Theparticlepassesthroughtheoriginwhens(t ) = t
4
−18t
2
= t
2
(t
2
−18) = 0. Thishappenswhent = 0
secondsandwhent = 3

2≈ 4.24seconds.Withs(t ) = t
4
−18t
2
,itfollowsthatv(t ) = s

(t ) = 4t
3
−36t = 4t (t
2
−9).
Theparticleisthereforeinstantaneouslymotionlesswhent = 0secondsandwhent = 3seconds.
Plot thepositionof theparticleinExercise23. What isthefarthest distancetotheleft of theoriginattainedby
theparticle?
25. A bullet is ﬁredintheair vertically fromgroundlevel withaninitial velocity 200m/s. Findthebullet’s maximum
velocityandmaximumheight.
solution Weemploy Galileo’sformula, s(t ) = s
0
+ v
0
t −
1
2
gt
2
= 200t − 4.9t
2
, wherethetimet isinseconds(s)
andtheheight s isinmeters(m). Thevelocityisv(t ) = 200−9.8t . Themaximumvelocityof 200m/soccursat t = 0.
Thisistheinitial velocity. Thebullet reachesitsmaximumheight whenv(t ) = 200−9.8t = 0; i.e., whent ≈ 20.41s.
At thispoint, theheight is2040.82m.
Findthevelocityof anobject droppedfromaheight of 300mat themoment it hitstheground.
27. A ball tossedintheair verticallyfromgroundlevel returnstoearth4slater. Findtheinitial velocityandmaximum
height of theball.
solution Galileo’sformulagivess(t ) = s
0
+ v
0
t −
1
2
gt
2
= v
0
t − 4.9t
2
, wherethetimet isinseconds(s) andthe
height s isinmeters(m). Whentheball hitsthegroundafter 4secondsitsheight is0. Solve0= s(4) = 4v
0
− 4.9(4)
2
toobtainv
0
= 19.6m/s. Theball reachesitsmaximumheight whens

(t ) = 0, that is, when19.6− 9.8t = 0, or t = 2
s. At thistime, t = 2s,
s(2) = 0+19.6(2) −
1
2
(9.8)(4) = 19.6m.
Oliviaisgazingout awindowfromthetenthﬂoor of abuildingwhenabucket (droppedby awindowwasher)
passesby. Shenotesthatithitstheground1.5slater. Determinetheﬂoor fromwhichthebucketwasdroppedif each
ﬂoor is5mhighandthewindowisinthemiddleof thetenthﬂoor. Neglect air friction.
29. Showthat for anobject fallingaccordingtoGalileo’sformula, theaveragevelocityover anytimeinterval [t
1
, t
2
] is
equal totheaverageof theinstantaneousvelocitiesat t
1
andt
2
.
solution Thesimplest waytoproceedistocomputebothvaluesandshowthat theyareequal. Theaveragevelocity
over [t
1
, t
2
] is
s(t
2
) −s(t
1
)
t
2
−t
1
=
(s
0
+v
0
t
2

1
2
gt
2
2
) −(s
0
+v
0
t
1

1
2
gt
2
1
)
t
2
−t
1
=
v
0
(t
2
−t
1
) +
g
2
(t
2
2
−t
1
2
)
t
2
−t
1
=
v
0
(t
2
−t
1
)
t
2
−t
1

g
2
(t
2
+t
1
) = v
0

g
2
(t
2
+t
1
)
Whereastheaverageof theinstantaneousvelocitiesat thebeginningandendof [t
1
, t
2
] is
s

(t
1
) +s

(t
2
)
2
=
1
2
_
(v
0
−gt
1
) +(v
0
−gt
2
)
_
=
1
2
(2v
0
) −
g
2
(t
2
+t
1
) = v
0

g
2
(t
2
+t
1
).
Thetwoquantitiesarethesame.
An object falls under theinﬂuenceof gravity near theearth’s surface. Which of thefollowingstatements is
true? Explain.
(a) Distancetraveledincreasesbyequal amountsinequal timeintervals.
(b) Velocityincreasesbyequal amountsinequal timeintervals.
(c) Thederivativeof velocityincreaseswithtime.
31. ByFaraday’sLaw, if aconductingwireof length metersmovesat velocityv m/sperpendicular toamagneticﬁeld
of strengthB (inteslas), avoltageof sizeV = −Bv isinducedinthewire. Assumethat B = 2and = 0.5.
(a) CalculatedV/dv.
(b) Findtherateof changeof V withrespect totimet if v = 4t +9.
solution
(a) Assumingthat B = 2andl = 0.5, V = −2(0.5)v = −v. Therefore,
dV
dv
= −1.
(b) If v = 4t +9, thenV = −2(0.5)(4t +9) = −(4t +9). Therefore,
dV
dt
= −4.
ThevoltageV, current I, andresistanceR inacircuit arerelatedby Ohm’sLaw: V = IR, wheretheunitsare
volts, amperes, andohms. Assumethat voltageisconstant withV = 12volts. Calculate(specifyingunits):
(a) Theaveragerateof changeof I withrespect toR for theinterval fromR = 8toR = 8.1
(b) Therateof changeof I withrespect toR whenR = 8
(c) Therateof changeof R withrespect toI whenI = 1.5
33. Ethanﬁnds that withh hours of tutoring, heis abletoanswer correctly S(h) percent of theproblems ona
mathexam. Whichwouldyouexpect tobelarger: S

(3) or S

(30)? Explain.
solution Onepossiblegraphof S(h) is shownintheﬁgurebelowontheleft. This graphindicates that intheearly
hoursof workingwiththetutor, Ethanmakesrapidprogressinlearningthematerial but eventuallyapproacheseither the
limit of his ability to learnthematerial or themaximumpossiblescoreontheexam. Inthis scenario, S

(3) wouldbe
larger thanS

(30).
June 8, 2011 LTSV SSM Second Pass
S E C T I ON 3.4 Rates of Change 123
An alternativegraph of S(h) is shown belowon theright. Here, in theearly hours of working with thetutor little
progressismade(perhapsthetutor isassessinghowmuchEthanalreadyknows, hislearningstyle, hispersonality, etc.).
Thisisfollowedbyaperiodof rapidimprovement andﬁnallyalevelingoff asEthanreacheshismaximumscore. Inthis
scenario, S

(3) andS

(30) might beroughlyequal.
Hours of tutoring
P
e
r
c
e
n
t
a
g
e

c
o
r
r
e
c
t
Hours of tutoring
P
e
r
c
e
n
t
a
g
e

c
o
r
r
e
c
t
Supposeθ(t ) measurestheanglebetweenaclock’sminuteandhour hands. What isθ

(t ) at 3o’clock?
35. Todeterminedrugdosages, doctorsestimateaperson’sbodysurfacearea(BSA) (inmeterssquared) usingtheformula
BSA =

hm/60, whereh istheheight incentimetersandmthemassinkilograms. Calculatetherateof changeof BSA
withrespect to mass for apersonof constant height h = 180. What is this rateat m = 70andm = 80? Express your
result inthecorrect units. DoesBSA increasemorerapidlywithrespect tomassat lower or higher bodymass?
solution Assumingconstant height h = 180cm, let f (m) =

hm/60=

5
10
m betheformulafor bodysurfacearea
intermsof weight. Therateof changeof BSA withrespect tomassis
f

(m) =

5
10
_
1
2
m
−1/2
_
=

5
20

m
.
If m = 70kg, thisis
f

(70) =

5
20

70
=

14
280
≈ 0.0133631
m
2
kg
.
If m = 80kg,
f

(80) =

5
20

80
=
1
20

16
=
1
80
m
2
kg
.
Becausetherateof changeof BSA dependson1/

m, it isclear that BSA increasesmorerapidlyat lower bodymass.
TheatmosphericCO
2
level A(t ) at MaunaLoa, Hawaii at timet (inpartsper millionbyvolume) isrecordedby
theScrippsInstitutionof Oceanography. Thevaluesfor themonthsJ anuary–December 2007were
382.45, 383.68, 384.23, 386.26, 386.39, 385.87,
384.39, 381.78, 380.73, 380.81, 382.33, 383.69
(a) Assumingthatthemeasurementsweremadeontheﬁrstof eachmonth, estimateA

(t ) onthe15thof themonths
J anuary–November.
(b) InwhichmonthsdidA

(t ) takeonitslargest andsmallest values?
(c) InwhichmonthwastheCO
2
level most nearlyconstant?
37. Thetangent linestothegraphof f (x) = x
2
growsteeper asx increases. At what ratedotheslopesof thetangent
linesincrease?
solution Let f (x) = x
2
. Theslopes s of thetangent lines aregivenby s = f

(x) = 2x. Therateat whichthese
slopesareincreasingisds/dx = 2.
Figure13shows theheight y of amass oscillatingat theendof aspring. throughonecycleof theoscillation.
Sketchthegraphof velocityasafunctionof time.
In Exercises 39–46, use Eq. (3) to estimate the unit change.
39. Estimate

2−

1and

101−

100. Compareyour estimateswiththeactual values.
solution Let f (x) =

x = x
1/2
. Thenf

(x) =
1
2
(x
−1/2
). Weareusingthederivativetoestimatetheaveragerate
of change. That is,

x +h −

x
h
≈ f

(x),
sothat

x +h −

x ≈ hf

(x).
Thus,

2−

1 ≈ 1f

(1) =
1
2
(1) =
1
2
. Theactual value, to six decimal places, is 0.414214. Also,

101−

100 ≈
1f

(100) =
1
2
_
1
10
_
= 0.05. Theactual value, tosixdecimal places, is0.0498756.
Estimatef (4) −f (3) if f

(x) = 2
−x
. Thenestimatef (4), assumingthat f (3) = 12.
41. Let F(s) = 1.1s + 0.05s
2
bethestoppingdistanceas inExample3. CalculateF(65) andestimatetheincreasein
stoppingdistanceif speedisincreasedfrom65to66mph. Compareyour estimatewiththeactual increase.
solution Let F(s) = 1.1s +.05s
2
beasinExample3. F

(s) = 1.1+0.1s.
• ThenF(65) = 282.75ft andF

(65) = 7.6ft/mph.
• F

(65) ≈ F(66) − F(65) is approximately equal tothechangeinstoppingdistanceper 1mphincreaseinspeed
whentravelingat 65mph. Increasingspeedfrom65to66thereforeincreasesstoppingdistanceby approximately
7.6ft.
• The actual increase in stopping distance when speed increases from65 mph to 66 mph is F(66) − F(65) =
290.4−282.75= 7.65feet, whichdiffersbylessthanonepercent fromtheestimatefoundusingthederivative.
June 8, 2011 LTSV SSM Second Pass
124 C HA P T E R 3 DIFFERENTIATION
AccordingtoKleiber’sLaw, themetabolicrateP (inkilocaloriesper day) andbodymassm (inkilograms) of an
animal arerelatedbyathree-quarter-power law P = 73.3m
3/4
. Estimatetheincreaseinmetabolicratewhenbody
massincreasesfrom60to61kg.
43. Thedollarcostof producingx bagelsisC(x) = 300+0.25x −0.5(x/1000)
3
. Determinethecostof producing2000
bagelsandestimatethecost of the2001st bagel. Compareyour estimatewiththeactual cost of the2001st bagel.
solution Expandingthepower of 3yields
C(x) = 300+0.25x −5×10
−10
x
3
.
Thisallowsustoget thederivativeC

(x) = 0.25−1.5×10
−9
x
2
. Thecost of producing2000bagelsis
C(2000) = 300+0.25(2000) −0.5(2000/1000)
3
= 796
dollars. The cost of the 2001st bagel is, by deﬁnition, C(2001) − C(2000). By the derivative estimate, C(2001) −
C(2000) ≈ C

(2000)(1), sothecost of the2001st bagel isapproximately
C

(2000) = 0.25−1.5×10
−9
(2000
2
) = \$0.244.
C(2001) = 796.244, sotheexact cost of the2001st bagel isindistinguishablefromtheestimatedcost. Thefunctionis
verynearlylinear at thispoint.
Supposethedollar cost of producingx videocamerasisC(x) = 500x −0.003x
2
+10
−8
x
3
.
(a) Estimatethemarginal costatproductionlevel x = 5000andcompareitwiththeactual costC(5001) −C(5000).
(b) Comparethemarginal cost at x = 5000withtheaveragecost per camera, deﬁnedasC(x)/x.
45. Demandfor acommoditygenerallydecreasesasthepriceisraised. Supposethat thedemandfor oil (per capitaper
year) is D(p) = 900/p barrels, wherep is thedollar priceper barrel. Findthedemandwhenp = \$40. Estimatethe
decreaseindemandif p risesto\$41andtheincreaseif p declinesto\$39.
solution D(p) = 900p
−1
, soD

(p) = −900p
−2
. Whenthepriceis\$40abarrel, theper capitademandisD(40) =
22.5 barrels per year. With an increase in price from\$40 to \$41 a barrel, the change in demand D(41) − D(40) is
approximatelyD

(40) = −900(40
−2
) = −0.5625barrelsayear. Withadecreaseinpricefrom\$40to\$39abarrel, the
changeindemandD(39) −D(40) isapproximately−D

(40) = +0.5625. Anincreaseinoil pricesof adollar leadstoa
decreaseindemandof 0.5625barrelsayear, andadecreaseof adollar leadstoanincrease indemandof 0.5625barrels
ayear.
Thereproduction ratef of thefruit ﬂy Drosophila melanogaster, grown in bottles in a laboratory, decreases
withthenumber p of ﬂies inthebottle. A researcher has foundthenumber of offspringper femaleper day to be
approximatelyf (p) = (34−0.612p)p
−0.658
.
(a) Calculatef (15) andf

(15).
(b) Estimatethedecreaseindailyoffspringper femalewhenp isincreasedfrom15to16. Isthisestimatelarger or
smaller thantheactual valuef (16) −f (15)?
(c) Plot f (p) for 5 ≤ p ≤ 25andverify that f (p) is adecreasingfunctionof p. Do youexpect f

(p) to be
positiveor negative? Plot f

(p) andconﬁrmyour expectation.
47. AccordingtoStevens’ Lawinpsychology, theperceivedmagnitudeof astimulusisproportional (approxi-
mately) toapower of theactual intensityI of thestimulus. Experimentsshowthat theperceived brightness B of alight
satisﬁesB = kI
2/3
, whereI isthelightintensity, whereastheperceived heaviness H of aweightW satisﬁesH = kW
3/2
(k isaconstant that isdifferent inthetwocases). ComputedB/dI anddH/dW andstatewhether theyareincreasingor
decreasingfunctions. Thenexplainthefollowingstatements:
(a) A one-unit increaseinlight intensityisfelt morestronglywhenI issmall thanwhenI islarge.
(b) Addinganother poundtoaloadW isfelt morestronglywhenW islargethanwhenW issmall.
solution
(a) dB/dI =
2k
3
I
−1/3
=
2k
3I
1/3
.
As I increases, dB/dI shrinks, so that the rate of change of perceived intensity decreases as the actual intensity
increases. Increasedlight intensityhasadiminished return inperceivedintensity. A sketchof B against I isshown: See
that theheight of thegraphincreasesmoreslowlyasyoumovetotheright.
(b) dH/dW =
3k
2
W
1/2
. As W increases, dH/dW increases as well, so that therateof changeof perceived weight
increasesasweight increases. A sketchof H against W isshown: Seethat thegraphbecomessteeper asyoumovetothe
right.
June 8, 2011 LTSV SSM Second Pass
S E C T I ON 3.4 Rates of Change 125
LetM(t ) bethemass(inkilograms) of aplantasafunctionof time(inyears). RecentstudiesbyNiklasandEnquist
havesuggestedthataremarkablywiderangeof plants(fromalgaeandgrasstopalmtrees)obeyathree-quarter-power
growth law—that is, dM/dt = CM
3/4
for someconstant C.
(a) If atreehasagrowthrateof 6kg/yr whenM = 100kg, what isitsgrowthratewhenM = 125kg?
(b) If M = 0.5kg, howmuchmoremassmust theplant acquiretodoubleitsgrowthrate?
Further Insights and Challenges
Exercises 49–51: The Lorenz curve y = F(r) is used by economists to study income distribution in a given country (see
Figure 14). By deﬁnition, F(r) is the fraction of the total income that goes to the bottomrth part of the population, where
0≤ r ≤ 1. For example, if F(0.4) = 0.245, then the bottom 40%of households receive 24.5%of the total income. Note
that F(0) = 0and F(1) = 1.
0.2 0.4 0.6 0.8 1.0
0.2
0.4
0.6
F(r)
0.2 0.4 0.6 0.8 1.0
0.2
0.4
0.6
0.8
1.0
F(r)
P
(A) Lorenz curvefor Sweden in 2004 (B) Two Lorenz curves: Thetangent
lines at P and Q haveslope1.
Q
0.2 0.4 0.6 0.8 1.0
0.2
0.4
0.6
0.8
1.0
r
F(r)
0.2 0.4 0.6 0.8 1.0
0.2
0.4
0.6
0.8
1.0
r
F(r)
L
2
L
1
P
(A) Lorenz curvefor Sweden in 2004 (B) Two Lorenz curves: Thetangent
lines at P and Q haveslope1.
Q
FIGURE 14
49. Our goal is to ﬁndaninterpretationfor F

(r). Theaverageincomefor agroupof households is thetotal
incomegoingtothegroupdividedbythenumber of householdsinthegroup. Thenational averageincomeisA = T/N,
whereN isthetotal number of householdsandT isthetotal incomeearnedbytheentirepopulation.
(a) Showthat theaverageincomeamonghouseholdsinthebottomrthpart isequal to(F(r)/r)A.
(b) Showmoregenerallythat theaverageincomeof householdsbelongingtoaninterval [r, r +r] isequal to
_
F(r +r) −F(r)
r
_
A
(c) Let 0≤ r ≤ 1. A householdbelongstothe100rthpercentileif itsincomeisgreater thanor equal totheincomeof
100r %of all households. Passtothelimit asr → 0in(b) toderivethefollowinginterpretation: A householdinthe
100rthpercentilehasincomeF

(r)A. Inparticular, ahouseholdinthe100rthpercentilereceivesmorethanthenational
averageif F

(r) > 1andlessif F

(r) < 1.
(d) For theLorenz curvesL
1
andL
2
inFigure14(B), what percentageof householdshaveabove-averageincome?
solution
(a) Thetotal incomeamonghouseholdsinthebottomrthpart isF(r)T andtherearerN householdsinthispart of the
population. Thus, theaverageincomeamonghouseholdsinthebottomrthpart isequal to
F(r)T
rN
=
F(r)
r
·
T
N
=
F(r)
r
A.
(b) Consider theinterval [r, r + r]. Thetotal incomeamonghouseholdsbetweenthebottomrthpart andthebottom
r +r-thpart isF(r +r)T −F(r)T . Moreover, thenumber of householdscoveredbythisinterval is(r +r)N −
rN = rN. Thus, theaverageincomeof householdsbelongingtoaninterval [r, r +r] isequal to
F(r +r)T −F(r)T
rN
=
F(r +r) −F(r)
r
·
T
N
=
F(r +r) −F(r)
r
A.
(c) Taketheresult frompart (b) andlet r → 0. Because
lim
r→0
F(r +r) −F(r)
r
= F

(r),
weﬁndthat ahouseholdinthe100rthpercentilehasincomeF

(r)A.
(d) ThepointP inFigure14(B) hasanr-coordinateof 0.6, whilethepointQhasanr-coordinateof roughly0.75. Thus,
oncurveL
1
, 40%of householdshaveF

(r) > 1andthereforehaveabove-averageincome. OncurveL
2
, roughly25%
of householdshaveabove-averageincome.
Thefollowingtableprovidesvaluesof F(r) for Swedenin2004. Assumethat thenational averageincomewas
A = 30,000euros.
r 0 0.2 0.4 0.6 0.8 1
F(r) 0 0.01 0.245 0.423 0.642 1
(a) What wastheaverageincomeinthelowest 40%of households?
(b) Showthat theaverageincomeof thehouseholdsbelongingtotheinterval [0.4, 0.6] was26,700euros.
(c) Estimate F

(0.5). Estimate the income of households in the 50th percentile? Was it greater or less than the
ti l ?
51. UseExercise49(c) toprove:
(a) F

(r) isanincreasingfunctionof r.
(b) Incomeisdistributedequally(all householdshavethesameincome) if andonlyif F(r) = r for 0≤ r ≤ 1.
June 8, 2011 LTSV SSM Second Pass
126 C HA P T E R 3 DIFFERENTIATION
solution
(a) Recall fromExercise49(c) that F

(r)A istheincomeof ahouseholdinthe100r-thpercentile. Suppose0≤ r
1
<
r
2
≤ 1. Becauser
2
> r
1
, ahouseholdinthe100r
2
-thpercentilemust haveincomeat least aslargeasahouseholdinthe
100r
1
-thpercentile. Thus, F

(r
2
)A ≥ F

(r
1
)A, or F

(r
2
) ≥ F

(r
1
). ThisimpliesF

(r) isanincreasingfunctionof r.
(b) If F(r) = r for 0 ≤ r ≤ 1, then F

(r) = 1andhouseholds in all percentiles haveincomeequal to thenational
average; that is, incomeisdistributedequally. Alternately, if incomeisdistributedequally(all householdshavethesame
income), thenF

(r) = 1for 0 ≤ r ≤ 1. Thus, F must bealinear functioninr withslope1. Moreover, thecondition
F(0) = 0requirestheF intercept of thelinetobe0. Hence, F(r) = 1· r +0= r.
Studiesof Internetusageshowthatwebsitepopularityisdescribedquitewell byZipf’sLaw, accordingtowhich
thenthmost popular websitereceivesroughlythefraction1/n of all visits. Supposethat onaparticular day, thenth
most popular sitehadapproximatelyV(n) = 10
6
/n visitors(for n ≤ 15,000).
(a) Verify that thetop50websites receivednearly 45%of thevisits. Hint: Let T (N) denotethesumof V(n) for
1≤ n ≤ N. Useacomputer algebrasystemtocomputeT (45) andT (15,000).
(b) Verify, by numerical experimentation, that whenEq. (3) is usedtoestimateV(n + 1) − V(n), theerror inthe
estimatedecreases as n grows larger. Find (again, by experimentation) an N such that theerror is at most 10for
n ≥ N.
(c) UsingEq. (3), showthat for n ≥ 100, thenth websitereceivedat most 100morevisitors than the(n + 1)st
website.
In Exercises 53 and 54, the average cost per unit at production level x is deﬁned as C
avg
(x) = C(x)/x, where C(x) is
the cost function. Average cost is a measure of the efﬁciency of the production process.
53. Showthat C
avg
(x) isequal totheslopeof thelinethroughtheoriginandthepoint (x, C(x)) onthegraphof C(x).
Usingthisinterpretation, determinewhether averagecost or marginal cost isgreater at pointsA, B, C, D inFigure15.
C
x
A
B C
D
FIGURE 15 Graphof C(x).
solution Bydeﬁnition, theslopeof thelinethroughtheoriginand(x, C(x)), that is, between(0, 0) and(x, C(x)) is
C(x) −0
x −0
=
C(x)
x
= C
av
.
At point A, averagecost isgreater thanmarginal cost, asthelinefromtheorigintoA issteeper thanthecurveat this
point(weseethisbecausetheline, tracingfromtheorigin, crossesthecurvefrombelow). Atpoint B, theaveragecostis
still greater thanthemarginal cost. At thepoint C, theaveragecost andthemarginal cost arenearlythesame, sincethe
tangent lineandthelinefromtheoriginarenearly thesame. ThelinefromtheorigintoD crossesthecost curvefrom
above, andsoislesssteepthanthetangent linetothecurveat D; theaveragecost at thispoint islessthanthemarginal
cost.
Thecost indollarsof producingalarmclocksisC(x) = 50x
3
− 750x
2
+ 3740x + 3750wherex isinunitsof
1000.
(a) Calculatetheaveragecost at x = 4, 6, 8, and10.
(b) Usethegraphical interpretationof averagecost toﬁndtheproductionlevel x
0
at whichaveragecost islowest.
What istherelationbetweenaveragecost andmarginal cost at x
0
(seeFigure16)?
3.5 Higher Derivatives
Preliminary Questions
1. OnSeptember 4, 2003, theWall Street Journal printedtheheadline“Stocks Go Higher, ThoughthePaceof Their
GainsSlows.”Rephrasethisheadlineasastatementabouttheﬁrstandsecondtimederivativesof stockpricesandsketch
apossiblegraph.
solution Becausestocks aregoing higher, stock prices areincreasing and theﬁrst derivativeof stock prices must
thereforebepositive. Ontheother hand, becausethepaceof gainsisslowing, thesecondderivativeof stockpricesmust
benegative.
Stock
price
Time
2. Trueor false? Thethirdderivativeof positionwithrespect to timeis zero for anobject fallingto earthunder the
inﬂuenceof gravity. Explain.
solution Thisstatementistrue. Theaccelerationof anobjectfallingtoearthundertheinﬂuenceof gravityisconstant;
hence, thesecondderivativeof positionwithrespecttotimeisconstant. Becausethethirdderivativeisjustthederivative
of thesecondderivativeandthederivativeof aconstant iszero, it followsthat thethirdderivativeiszero.
3. Whichtypeof polynomial satisﬁesf

(x) = 0for all x?
solution Thethirdderivativeof all quadraticpolynomials(polynomialsof theformax
2
+bx +c for someconstants
a, b andc) isequal to0for all x.
June 8, 2011 LTSV SSM Second Pass
S E C T I ON 3.5 Higher Derivatives 127
4. What isthesixthderivativeof f (x) = x
6
?
solution Thesixthderivativeof f (x) = x
6
is6! = 720.
Exercises
In Exercises 1–16, calculate y

and y

.
1. y = 14x
2
solution Let y = 14x
2
. Theny

= 28x, y

= 28, andy

= 0.
y = 7−2x
3. y = x
4
−25x
2
+2x
solution Let y = x
4
−25x
2
+2x. Theny

= 4x
3
−50x +2, y

= 12x
2
−50, andy

= 24x.
y = 4t
3
−9t
2
+7
5. y =
4
3
πr
3
solution Let y =
4
3
πr
3
. Theny

= 4πr
2
, y

= 8πr, andy

= 8π.
y =

x
7. y = 20t
4/5
−6t
2/3
solution Lety = 20t
4/5
−6t
2/3
. Theny

= 16t
−1/5
−4t
−1/3
, y

= −
16
5
t
−6/5
+
4
3
t
−4/3
, andy

=
96
25
t
−11/15

16
9
t
−7/3
.
y = x
−9/5 9. y = z −
4
z
solution Let y = z −4z
−1
. Theny

= 1+4z
−2
, y

= −8z
−3
, andy

= 24z
−4
.
y = 5t
−3
+7t
−8/3
11. y = θ
2
(2θ +7)
solution Let y = θ
2
(2θ +7) = 2θ
3
+7θ
2
. Theny

= 6θ
2
+14θ, y

= 12θ +14, andy

= 12.
y = (x
2
+x)(x
3
+1)
13. y =
x −4
x
solution Let y =
x−4
x
= 1−4x
−1
. Theny

= 4x
−2
, y

= −8x
−3
, andy

= 24x
−4
.
y =
1
1−x
15. y = s
−1/2
(s +1)
solution Expandinggivesy = s
1/2
+s
−1/2
. Then
y

=
1
2
s
−1/2

1
2
s
−3/2
y

= −
1
4
s
−3/2
+
3
4
s
−5/2
y

=
3
8
s
−5/2

15
8
s
−7/2
y = (r
1/2
+r)(1−r)
In Exercises 17–25, calculate the derivative indicated.
17. f
(4)
(1), f (x) = x
4
solution Let f (x) = x
4
. Thenf

(x) = 4x
3
, f

(x) = 12x
2
, f

(x) = 24x, andf
(4)
(x) = 24. Thusf
(4)
(1) = 24.
g

(−1), g(t ) = −4t
−5
19.
d
2
y
dt
2
¸
¸
¸
¸
t =1
, y = 4t
−3
+3t
2
solution Let y = 4t
−3
+3t
2
. Then
dy
dt
= −12t
−4
+6t and
d
2
y
dt
2
= 48t
−5
+6. Hence
d
2
y
dt
2
¸
¸
¸
¸
¸
t =1
= 48(1)
−5
+6= 54.
d
4
f
dt
4
¸
¸
¸
¸
t =1
, f (t ) = 6t
9
−2t
5
21.
d
4
x
dt
4
¸
¸
¸
¸
t =16
, x = t
−3/4
solution Let x(t ) = t
−3/4
. Then
dx
dt
= −
3
4
t
−7/4
,
d
2
x
dt
2
=
21
16
t
−11/4
,
d
3
x
dt
3
= −
231
64
t
−15/4
, and
d
4
x
dt
4
=
3465
256
t
−19/4
.
Thus
d
4
x
dt
4
¸
¸
¸
¸
¸
t =16
=
3465
256
16
−19/4
=
3465
134,217,728
.
June 8, 2011 LTSV SSM Second Pass
128 C HA P T E R 3 DIFFERENTIATION
f

(4), f (t ) = 2t
2
−t
23. f

(−3), f (x) =
12
x
−x
3
solution Differentiatinggives
f

(x) = −12x
−2
−3x
2
f

(x) = 24x
−3
−6x
f

(x) = −72x
−4
−6
Thus, f

(−3) = −72· 3
−4
−6= −
62
9
.
f

(1), f (t ) =
t
t +1
25. h

(1), h(w) =
1

w +1
solution Differentiatinggives
h

(w) =

1
2
w
−1/2
(

w +1)
2
= −
1
2

w(

w +1)
2
h

(w) =
2

w
_
2(

w +1)
1
2
w
−1/2
_
+2·
1
2
w
−1/2
(

w +1)
2
4w(

w +1)
4
=
2(

w +1) +w
−1/2
(

w +1)
2
4w(

w +1)
4
sothat
h

(1) =
2(1+1) +1(1+1)
2
4· 1(1+1)
4
=
8
64
=
1
8
g

(1), g(s) =

s
s +1
27. Calculatey
(k)
(0) for 0≤ k ≤ 5, wherey = x
4
+ax
3
+bx
2
+cx +d (witha, b, c, d theconstants).
solution Applyingthepower, constant multiple, andsumrulesat eachstage, weget (notey
(0)
isy byconvention):
k y
(k)
0 x
4
+ax
3
+bx
2
+cx +d
1 4x
3
+3ax
2
+2bx +c
2 12x
2
+6ax +2b
3 24x +6a
4 24
5 0
fromwhichweget y
(0)
(0) = d, y
(1)
(0) = c, y
(2)
(0) = 2b, y
(3)
(0) = 6a, y
(4)
(0) = 24, andy
(5)
(0) = 0.
Whichof thefollowingsatisfyf
(k)
(x) = 0for all k ≥ 6?
(a) f (x) = 7x
4
+4+x
−1
(b) f (x) = x
3
−2
(c) f (x) =

x (d) f (x) = 1−x
6
(e) f (x) = x
9/5
(f) f (x) = 2x
2
+3x
5
29. Usetheresult inExample3toﬁnd
d
6
dx
6
x
−1
.
solution TheequationinExample3indicatesthat
d
6
dx
6
x
−1
= (−1)
6
6!x
−6−1
.
(−1)
6
= 1and6! = 6×5×4×3×2×1= 720, so
d
6
dx
6
x
−1
= 720x
−7
.
Calculatetheﬁrst ﬁvederivativesof f (x) =

x.
(a) Showthat f
(n)
(x) isamultipleof x
−n+1/2
.
(b) Showthat f
(n)
(x) alternatesinsignas(−1)
n−1
for n ≥ 1.
(c) Findaformulafor f
(n)
(x) for n ≥ 2. Hint: Verifythat thecoefﬁcient is±1· 3· 5· · ·
2n −3
2
n
.
In Exercises 31–36, ﬁnd a general formula for f
(n)
(x).
31. f (x) = x
−2
solution f

(x) = −2x
−3
, f

(x) = 6x
−4
, f

(x) = −24x
−5
, f
(4)
(x) = 5· 24x
−6
, . . . . Fromthiswecanconclude
that thenthderivativecanbewrittenasf
(n)
(x) = (−1)
n
(n +1)!x
−(n+2)
.
June 8, 2011 LTSV SSM Second Pass
S E C T I ON 3.5 Higher Derivatives 129
f (x) = (x +2)
−1
33. f (x) = x
−1/2
solution f

(x) =
−1
2
x
−3/2
. Wewill avoidsimplifyingnumeratorsanddenominatorstoﬁndthepattern:
f

(x) =
−3
2
−1
2
x
−5/2
= (−1)
2
3×1
2
2
x
−5/2
f

(x) = −
5
2
3×1
2
2
x
−7/2
= (−1)
3
5×3×1
2
3
x
−7/2
.
.
.
f
(n)
(x) = (−1)
n
(2n −1) ×(2n −3) ×. . . ×1
2
n
x
−(2n+1)/2
.
f (x) = x
−3/2 35. f (x) =
x +1
x
2
solution Let f (x) =
x +1
x
2
= x
−1
+x
−2
. Takingsuccessivederivativesgives
f

(x) = −x
−2
−2x
−3
f

(x) = 2x
−3
+6x
−4
f

(x) = −6x
−4
−24x
−5
f
(4)
(x) = 24x
−5
+120x
−6
.
.
.
f
(n)
(x) = (−1)
n
(n!x
−n−1
+(n +1)!x
−n−2
)
f (x) =
x −1

x
37. (a) Findtheaccelerationat timet = 5minof ahelicopter whoseheight iss(t ) = 300t −4t
3
m.
(b) Plot theaccelerationh

(t ) for 0≤ t ≤ 6. Howdoesthisgraphshowthat thehelicopter isslowingdownduringthis
timeinterval?
solution
(a) Lets(t ) = 300t −4t
3
, witht inminutesands inmeters. Thevelocityisv(t ) = s

(t ) = 300−12t
2
andacceleration
isa(t ) = s

(t ) = −24t . Thusa(5) = −120m/min
2
.
(b) Theaccelerationof thehelicopter for 0≤ t ≤ 6isshownintheﬁgurebelow. Astheaccelerationof thehelicopter is
negative, thevelocityof thehelicopter must bedecreasing. Becausethevelocityispositivefor 0≤ t ≤ 6, thehelicopter
isslowingdown.
−140
−120
−100
−80
−60
−40
−20
1 2 3 4 5 6
y
x
Findanequationof thetangent tothegraphof y = f

(x) at x = 3, wheref (x) = x
4
.
39. Figure5showsf , f

, andf

. Determinewhichiswhich.
(A) (B)
x
y
3 2 1
x
y
3 2 1
x
y
(C)
3 2 1
FIGURE 5
solution (a) f

(b) f

(c) f .
Thetangent lineto(c) ishorizontal at x = 1andx = 3, where(b) hasroots. Thetangent lineto(b) ishorizontal at
x = 2andx = 0, where(a) hasroots.
Thesecondderivativef

isshowninFigure6. Whichof (A) or (B) isthegraphof f andwhichisf

?
June 8, 2011 LTSV SSM Second Pass
130 C HA P T E R 3 DIFFERENTIATION
41. Figure7showsthegraphof thepositions of anobject asafunctionof timet . Determinetheintervalsonwhichthe
accelerationispositive.
Time
40 30 20 10
Position
FIGURE 7
solution Roughly fromtime10totime20andfromtime30totime40. Theaccelerationispositiveover thesame
intervalsover whichthegraphisbendingupward.
Findapolynomial f (x) that satisﬁestheequationxf

(x) +f (x) = x
2
.
43. Findall valuesof n suchthat y = x
n
satisﬁes
x
2
y

−2xy

= 4y
solution Wehavey

= nx
n−1
, y

= n(n −1)x
n−2
, sothat
x
2
y

−2xy

= x
2
(n(n −1)x
n−2
) −2xnx
n−1
= (n
2
−3n)x
n
= (n
2
−3n)y
Thustheequationissatisﬁedif andonlyif n
2
−3n = 4, sothat n
2
−3n −4= 0. Thishappensfor n = −1, 4.
Whichof thefollowingdescriptionscouldnot applytoFigure8? Explain.
(a) Graphof accelerationwhenvelocityisconstant
(b) Graphof velocitywhenaccelerationisconstant
(c) Graphof positionwhenaccelerationiszero
45. Accordingtoonemodel that takesintoaccount air resistance, theaccelerationa(t ) (inm/s
2
) of askydiver of mass
m infreefall satisﬁes
a(t ) = −9.8+
k
m
v(t )
2
wherev(t ) isvelocity(negativesincetheobjectisfalling) andk isaconstant. Supposethatm = 75kgandk = 14kg/m.
(a) What istheobject’svelocitywhena(t ) = −4.9?
(b) What istheobject’svelocitywhena(t ) = 0?Thisvelocityistheobject’sterminal velocity.
solution Solvinga(t ) = −9.8+
k
m
v(t )
2
for thevelocity andtakingintoaccount that thevelocity isnegativesince
theobject isfalling, weﬁnd
v(t ) = −
_
m
k
(a(t ) +9.8) = −
_
75
14
(a(t ) +9.8).
(a) Substitutinga(t ) = −4.9intotheaboveformulafor thevelocity, weﬁnd
v(t ) = −
_
75
14
(4.9) = −

26.25= −5.12m/s.
(b) Whena(t ) = 0,
v(t ) = −
_
75
14
(9.8) = −

52.5= −7.25m/s.
Accordingtoonemodel that attemptstoaccount for air resistance, thedistances(t ) (inmeters) traveledby a
fallingraindropsatisﬁes
d
2
s
dt
2
= g −
0.0005
D
_
ds
dt
_
2
whereD istheraindropdiameter andg = 9.8m/s
2
. Terminal velocity v
term
isdeﬁnedasthevelocity at whichthe
drophaszeroacceleration(onecanshowthat velocityapproachesv
term
astimeproceeds).
(a) Showthat v
term
=

2000gD.
(b) Findv
term
for dropsof diameter 10
−3
mand10
−4
m.
(c) Inthismodel, doraindropsacceleratemorerapidlyat higher or lower velocities?
47. Aservomotorcontrolsthevertical movementof adrill bitthatwill drill apatternof holesinsheetmetal.Themaximum
vertical speedof thedrill bit is4in./s, andwhiledrillingthehole, it must movenomorethan2.6in./stoavoidwarping
themetal. Duringacycle, thebit beginsandendsat rest, quickly approachesthesheet metal, andquickly returnstoits
initial positionafter theholeisdrilled. Sketchpossiblegraphsof thedrill bit’svertical velocity andacceleration. Label
thepoint wherethebit entersthesheet metal.
solution Therewill bemultiplecycles, eachof whichwill bemoreorlessidentical. Letv(t ) bethedownward vertical
velocity of thedrill bit, andlet a(t ) bethevertical acceleration. Fromthenarrative, weseethat v(t ) canbeno greater
than4andnogreater than2.6whiledrillingistakingplace. Duringeachcycle, v(t ) = 0initially, v(t ) goesto4quickly.
Whenthebit hitsthesheet metal, v(t ) goesdownto2.6quickly, at whichit staysuntil thesheet metal isdrilledthrough.
Asthedrill pullsout, itreachesmaximumnon-drillingupwardspeed(v(t ) = −4) quickly, andmaintainsthisspeeduntil
it returnstorest. A possibleplot follows:
−2
−4
4
2
2 1.5 1 0.5
x
y
Metal
June 8, 2011 LTSV SSM Second Pass
S E C T I ON 3.5 Higher Derivatives 131
A graphof theaccelerationisextractedfromthisgraph:
−20
−40
40
20
2 1.5
1 0.5
x
y
Metal
In Exercises 48 and 49, refer to the following. In a 1997 study, Boardman and Lave related the trafﬁc speed S on a
two-lane road to trafﬁc density Q (number of cars per mile of road) by the formula
S = 2882Q
−1
−0.052Q+31.73
for 60≤ Q ≤ 400(Figure 9).
Density Q
400 300 200 100
Speed S
(mph)
10
20
30
40
50
60
70
FIGURE 9 Speedasafunctionof trafﬁcdensity.
CalculatedS/dQandd
2
S/dQ
2
.
49. (a) Explainintuitivelywhyweshouldexpect that dS/dQ < 0.
(b) Showthat d
2
S/dQ
2
> 0. Thenusethefact that dS/dQ < 0andd
2
S/dQ
2
> 0tojustifythefollowingstatement:
A one-unit increase in trafﬁc density slows down trafﬁc more when Q is small than when Q is large.
(c) Plot dS/dQ. Whichpropertyof thisgraphshowsthat d
2
S/dQ
2
> 0?
solution
(a) Trafﬁcspeedmust bereducedwhentheroadgetsmorecrowdedsoweexpect dS/dQtobenegative. Thisisindeed
thecasesincedS/dQ = −0.052−2882/Q
2
< 0.
(b) Thedecreasein speed dueto aone-unit increasein density is approximately dS/dQ (anegativenumber). Since
d
2
S/dQ
2
= 5764Q
−3
> 0ispositive, thistellsusthatdS/dQgetslargerasQincreases—andanegativenumberwhich
gets larger is gettingcloser tozero. Sothedecreaseinspeedis smaller whenQ is larger, that is, aone-unit increasein
trafﬁcdensityhasasmaller effect whenQislarge.
(c) dS/dQisplottedbelow. Thefact that thisgraphisincreasingshowsthat d
2
S/dQ
2
> 0.
x
y
−0.2
−0.4
−0.6
−0.8
−1
−1.2
400 300 100 200
Useacomputer algebrasystemtocomputef
(k)
(x) for k = 1, 2, 3for thefollowingfunctions.
(a) f (x) = (1+x
3
)
5/3
(b) f (x) =
1−x
4
1−5x −6x
2
51. Let f (x) =
x +2
x −1
. Useacomputer algebrasystemtocomputethef
(k)
(x) for 1≤ k ≤ 4. Canyouﬁnda
general formulafor f
(k)
(x)?
solution Let f (x) =
x +2
x −1
. Usingacomputer algebrasystem,
f

(x) = −
3
(x −1)
2
= (−1)
1
3· 1
(x −1)
1+1
;
f

(x) =
6
(x −1)
3
= (−1)
2
3· 2· 1
(x −1)
2+1
;
f

(x) = −
18
(x −1)
4
= (−1)
3
3· 3!
(x −1)
3+1
; and
f
(4)
(x) =
72
(x −1)
5
= (−1)
4
3· 4!
(x −1)
4+1
.
June 8, 2011 LTSV SSM Second Pass
132 C HA P T E R 3 DIFFERENTIATION
Fromthepatternobservedabove, weconjecture
f
(k)
(x) = (−1)
k
3· k!
(x −1)
k+1
.
Further Insights and Challenges
Findthe100thderivativeof
p(x) = (x +x
5
+x
7
)
10
(1+x
2
)
11
(x
3
+x
5
+x
7
)
53. What isp
(99)
(x) for p(x) asinExercise52?
solution First notethat for anyinteger n ≤ 98,
d
99
dx
99
x
n
= 0.
Now, if weexpandp(x), weﬁnd
p(x) = x
99
+ termsof degreeat most 98;
therefore,
d
99
dx
99
p(x) =
d
99
dx
99
(x
99
+ termsof degreeat most 98) =
d
99
dx
99
x
99
Usinglogicsimilar tothat usedtocomputethederivativeinExample(3), wecompute:
d
99
dx
99
(x
99
) = 99×98×. . . 1,
sothat
d
99
dx
99
p(x) = 99!.
UsetheProductRuletwicetoﬁndaformulafor (fg)

intermsof f andg andtheir ﬁrstandsecondderivatives.
55. UsetheProduct Ruletoﬁndaformulafor (fg)

andcompareyour result withtheexpansionof (a +b)
3
. Thentry
toguessthegeneral formulafor (fg)
(n)
.
solution ContinuingfromExercise54, wehave
h

= f

g

+gf

+2(f

g

+g

f

) +fg

+g

f

= f

g +3f

g

+3f

g

+fg

Thebinomial theoremgives
(a +b)
3
= a
3
+3a
2
b +3ab
2
+b
3
= a
3
b
0
+3a
2
b
1
+3a
1
b
2
+a
0
b
3
andmoregenerally
(a +b)
n
=
n

k=0
_
n
k
_
a
n−k
b
k
,
wherethebinomial coefﬁcientsaregivenby
_
n
k
_
=
k(k −1) · · · (k −n +1)
n!
.
Accordingly, thegeneral formulafor (fg)
(n)
isgivenby
(fg)
(n)
=
n

k=0
_
n
k
_
f
(n−k)
g
(k)
,
wherep
(k)
isthekthderivativeof p (or p itself whenk = 0).
Compute
f (x) = lim
h→0
f (x +h) +f (x −h) −2f (x)
h
2
for thefollowingfunctions:
(a) f (x) = x (b) f (x) = x
2
(c) f (x) = x
3
Basedontheseexamples, what doyouthinkthelimit f represents?
3.6 Trigonometric Functions
Preliminary Questions
1. Determinethesign(+ or −) that yieldsthecorrect formulafor thefollowing:
(a)
d
dx
(sinx +cosx) = ±sinx ±cosx
(b)
d
dx
secx = ±secx tanx
(c)
d
dx
cotx = ±csc
2
x
June 8, 2011 LTSV SSM Second Pass
S E C T I ON 3.6 Trigonometric Functions 133
solution Thecorrect formulasare
(a)
d
dx
(sinx +cosx) = −sinx +cosx
(b)
d
dx
secx = secx tanx
(c)
d
dx
cotx = −csc
2
x
2. Whichof thefollowingfunctionscanbedifferentiatedusingtheruleswehavecoveredsofar?
(a) y = 3cosx cotx (b) y = cos(x
2
) (c) y = 2
x
sinx
solution
(a) 3cosx cotx isaproductof functionswhosederivativesareknown. Thisfunctioncanthereforebedifferentiatedusing
theProduct Rule.
(b) cos(x
2
) is acompositionof thefunctions cosx andx
2
. Wehavenot yet discussedhowto differentiatecomposite
functions.
(c) 2
x
sinx is a product of functions, but we do not yet know how to differentiate 2
x
, so we do not know how to
differentiatetheproduct.
3. Compute
d
dx
(sin
2
x +cos
2
x) without usingthederivativeformulasfor sinx andcosx.
solution Recall that sin
2
x +cos
2
x = 1for all x. Thus,
d
dx
(sin
2
x +cos
2
x) =
d
dx
1= 0.
4. Howistheadditionformulausedinderivingtheformula(sinx)

= cosx?
solution Thedifferencequotient for thefunctionsinx involvestheexpressionsin(x +h). Theadditionformulafor
thesinefunctionisusedtoexpandthisexpressionassin(x +h) = sinx cosh +sinhcosx.
Exercises
In Exercises 1–4, ﬁnd an equation of the tangent line at the point indicated.
1. y = sinx, x =
π
4
solution Let f (x) = sinx. Thenf

(x) = cosx andtheequationof thetangent lineis
y = f

_
π
4
_ _
x −
π
4
_
+f
_
π
4
_
=

2
2
_
x −
π
4
_
+

2
2
=

2
2
x +

2
2
_
1−
π
4
_
.
y = cosx, x =
π
3
3. y = tanx, x =
π
4
solution Let f (x) = tanx. Thenf

(x) = sec
2
x andtheequationof thetangent lineis
y = f

_
π
4
_ _
x −
π
4
_
+f
_
π
4
_
= 2
_
x −
π
4
_
+1= 2x +1−
π
2
.
y = secx, x =
π
6
In Exercises 5–24, compute the derivative.
5. f (x) = sinx cosx
solution Let f (x) = sinx cosx. Then
f

(x) = sinx(−sinx) +cosx(cosx) = −sin
2
x +cos
2
x.
f (x) = x
2
cosx
7. f (x) = sin
2
x
solution Let f (x) = sin
2
x = sinx sinx. Then
f

(x) = sinx(cosx) +sinx(cosx) = 2sinx cosx.
f (x) = 9secx +12cotx
9. H(t ) = sint sec
2
t
solution Let H(t ) = sint sec
2
t . Then
H

(t ) = sint
d
dt
(sect · sect ) +sec
2
t (cost )
= sint (sect sect tant +sect sect tant ) +sect
= 2sint sec
2
t tant +sect.
June 8, 2011 LTSV SSM Second Pass
134 C HA P T E R 3 DIFFERENTIATION
h(t ) = 9csct +t cott
11. f (θ) = tanθ secθ
solution Let f (θ) = tanθ secθ. Then
f

(θ) = tanθ secθ tanθ +secθ sec
2
θ = secθ tan
2
θ +sec
3
θ =
_
tan
2
θ +sec
2
θ
_
secθ.
k(θ) = θ
2
sin
2
θ
13. f (x) = (2x
4
−4x
−1
) secx
solution Let f (x) = (2x
4
−4x
−1
) secx. Then
f

(x) = (2x
4
−4x
−1
) secx tanx +secx(8x
3
+4x
−2
).
f (z) = z tanz 15. y =
secθ
θ
solution Let y =
secθ
θ
. Then
y

=
θ secθ tanθ −secθ
θ
2
.
G(z) =
1
tanz −cotz
17. R(y) =
3cosy −4
siny
solution Let R(y) =
3cosy −4
siny
. Then
R

(y) =
siny(−3siny) −(3cosy −4)(cosy)
sin
2
y
=
4cosy −3(sin
2
y +cos
2
y)
sin
2
y
=
4cosy −3
sin
2
y
.
f (x) =
x
sinx +2
19. f (x) =
1+tanx
1−tanx
solution Let f (x) =
1+tanx
1−tanx
. Then
f

(x) =
(1−tanx) sec
2
x −(1+tanx)
_
−sec
2
x
_
(1−tanx)
2
=
2sec
2
x
(1−tanx)
2
.
f (θ) = θ tanθ secθ 21. f (x) =
sinx +1
sinx −1
solution Wehave
f

(x) =
(sinx −1)(cosx) −(sinx +1)(cosx)
(sinx −1)
2
=
−2cosx
(sinx −1)
2
f (x) =
csc
2
t
t
23. R(θ) =
cosθ
4+cosθ
solution Differentiating, wehave
R

(θ) =
(4+cosθ)(−sinθ) −(cosθ)(−sinθ)
(4+cosθ)
2
= −
4sinθ
(4+cosθ)
2
g(z) =
cotz
3−3sinz
In Exercises 25–34, ﬁnd an equation of the tangent line at the point speciﬁed.
25. y = x
3
+cosx, x = 0
solution Let f (x) = x
3
+cosx. Thenf

(x) = 3x
2
−sinx andf

(0) = 0. Thetangent lineat x = 0is
y = f

(0)(x −0) +f (0) = 0(x) +1= 1.
y = tanθ, θ =
π
6
27. y = sinx +3cosx, x = 0
solution Let f (x) = sinx +3cosx. Thenf

(x) = cosx −3sinx andf

(0) = 1. Thetangent lineat x = 0is
y = f

(0)(x −0) +f (0) = x +3.
June 8, 2011 LTSV SSM Second Pass
S E C T I ON 3.6 Trigonometric Functions 135
y =
sint
1+cost
, t =
π
3
29. y = 2(sinθ +cosθ), θ =
π
3
solution Let f (θ) = 2(sinθ + cosθ). Thenf

(θ) = 2(cosθ − sinθ) andf

(
π
3
) = 1−

3. Thetangent lineat
x =
π
3
is
y = f

_
π
3
_ _
x −
π
3
_
+f
_
π
3
_
= (1−

3)
_
x −
π
3
_
+1+

3.
y = cscx −cotx, x =
π
4
31. y = (cott )(cost ), t =
π
3
solution Let f (t ) = (cott )(cost ). Then
f

(t ) = (−csc
2
t )(cost ) +(cott )(−sint ) = −cott csct −cost
For t =
π
3
, wehave
f (t ) = cot
π
3
cos
π
3
=

3
6
, f

(t ) = −cot
π
3
csc
π
3
−cos
π
3
= −
2
3

1
2
= −
7
6
sothat theequationof thetangent lineis
y = f

_
π
3
_ _
x −
π
3
_
+f
_
π
3
_
= −
7
6
x +

18
+

3
6
y = x cos
2
x, x =
π
4
33. y = x
2
(1−sinx), x =

2
solution Let f (x) = x
2
(1−sinx). Then
f

(x) = 2x(1−sinx) −x
2
cosx
For a =

2
, wehave
f (a) =

2
4
(1−(−1)) =

2
2
, f

(a) = 3π(1−(−1)) −

2
4
(0) = 6π
sothat theequationof thetangent lineis
y = f

(a)(x −a) +f (a) = 6π
_
x −

2
_
+

2
2
= 6πx −

2
2
y =
sinθ −cosθ
θ
, θ =
π
4
In Exercises 35–37, use Theorem 1 to verify the formula.
35.
d
dx
cotx = −csc
2
x
solution cotx =
cosx
sinx
. Usingthequotient ruleandthederivativeformulas, wecompute:
d
dx
cotx =
d
dx
cosx
sinx
=
sinx(−sinx) −cosx(cosx)
sin
2
x
=
−(sin
2
x +cos
2
x)
sin
2
x
=
−1
sin
2
x
= −csc
2
x.
d
dx
secx = secx tanx
37.
d
dx
cscx = −cscx cotx
solution Sincecscx =
1
sinx
, wecanapplythequotient ruleandthetwoknownderivativestoget:
d
dx
cscx =
d
dx
1
sinx
=
sinx(0) −1(cosx)
sin
2
x
=
−cosx
sin
2
x
= −
cosx
sinx
1
sinx
= −cotx cscx.
Showthat bothy = sinx andy = cosx satisfyy

= −y.
In Exercises 39–42, calculate the higher derivative.
39. f

(θ), f (θ) = θ sinθ
solution Let f (θ) = θ sinθ. Then
f

(θ) = θ cosθ +sinθ
f

(θ) = θ(−sinθ) +cosθ +cosθ = −θ sinθ +2cosθ.
June 8, 2011 LTSV SSM Second Pass
136 C HA P T E R 3 DIFFERENTIATION
d
2
dt
2
cos
2
t
41. y

, y

, y = tanx
solution Let y = tanx. Theny

= sec
2
x andbytheChainRule,
y

= =
d
dx
sec
2
x = 2(secx)(secx tanx) = 2sec
2
x tanx
y

= 2sec
2
x(sec
2
x) +(2sec
2
x tanx) tanx = 2sec
4
+4sec
4
x tan
2
x
y

, y

, y = t
2
sint
43. Calculatetheﬁrst ﬁvederivativesof f (x) = cosx. Thendeterminef
(8)
andf
(37)
.
solution Let f (x) = cosx.
• Thenf

(x) = −sinx, f

(x) = −cosx, f

(x) = sinx, f
(4)
(x) = cosx, andf
(5)
(x) = −sinx.
• Accordingly, thesuccessivederivativesof f cycleamong
{−sinx, −cosx, sinx, cosx}
inthat order. Since8isamultipleof 4, wehavef
(8)
(x) = cosx.
• Since36isamultipleof 4, wehavef
(36)
(x) = cosx. Therefore, f
(37)
(x) = −sinx.
Findy
(157)
, wherey = sinx.
45. Findthevaluesof x between0and2π wherethetangent linetothegraphof y = sinx cosx ishorizontal.
solution Let y = sinx cosx. Then
y

= (sinx)(−sinx) +(cosx)(cosx) = cos
2
x −sin
2
x.
Wheny

= 0, wehavesinx = ±cosx. Intheinterval [0, 2π], thisoccurswhenx =
π
4
,

4
,

4
,

4
.
Plot thegraphf (θ) = secθ + cscθ over [0, 2π] anddeterminethenumber of solutionstof

(θ) = 0inthis
interval graphically. Thencomputef

(θ) andﬁndthesolutions.
47. Let g(t ) = t −sint .
(a) Plot thegraphof g withagraphingutilityfor 0≤ t ≤ 4π.
(b) Showthat theslopeof thetangent lineisnonnegative. Verifythisonyour graph.
(c) For whichvaluesof t inthegivenrangeisthetangent linehorizontal?
solution Let g(t ) = t −sint .
(a) Hereisagraphof g over theinterval [0, 4π].
y
x
2 4 6 8 10 12
2
4
6
8
10
12
(b) Sinceg

(t ) = 1−cost ≥ 0for all t , theslopeof thetangent linetog isalwaysnonnegative.
(c) Intheinterval [0, 4π], thetangent lineishorizontal whent = 0, 2π, 4π.
Let f (x) = (sinx)/x for x = 0andf (0) = 1.
(a) Plot f (x) on[−3π, 3π].
(b) Showthat f

(c) = 0if c = tanc. Usethenumerical root ﬁnder onacomputer algebrasystemto ﬁndagood
approximationtothesmallest positive valuec
0
suchthat f

(c
0
) = 0.
(c) Verifythat thehorizontal liney = f (c
0
) istangent tothegraphof y = f (x) at x = c
0
byplottingthemonthe
sameset of axes.
49. Showthat notangent linetothegraphof f (x) = tanx haszeroslope. What istheleast slopeof atangent
line? J ustifybysketchingthegraphof (tanx)

.
solution Let f (x) = tanx. Thenf

(x) = sec
2
x =
1
cos
2
x
. Notethat f

(x) =
1
cos
2
x
hasnumerator 1; theequation
f

(x) = 0thereforehasnosolution. Becausethemaximumvalueof cos
2
x is1, theminimumvalueof f

(x) =
1
cos
2
x
is1. Hence, theleast slopefor atangent linetotanx is1. Hereisagraphof f

.
2 4 −2 −4
y
x
2
4
6
8
10
12
14
June 8, 2011 LTSV SSM Second Pass
S E C T I ON 3.6 Trigonometric Functions 137
Theheight at timet (inseconds) of amass, oscillatingat theendof aspring, iss(t ) = 300+ 40sint cm. Find
thevelocityandaccelerationat t =
π
3
s.
51. The horizontal range R of a projectile launched fromground level at an angle θ and initial velocity v
0
m/s is
R = (v
2
0
/9.8) sinθ cosθ. CalculatedR/dθ. If θ = 7π/24, will therangeincreaseor decreaseif theangleisincreased
slightly? Baseyour answer onthesignof thederivative.
solution Let R(θ) = (v
2
0
/9.8) sinθ cosθ.
dR

= R

(θ) = (v
2
0
/9.8)(−sin
2
θ +cos
2
θ).
If θ = 7π/24,
π
4
< θ <
π
2
, so| sinθ| > | cosθ|, anddR/dθ < 0(numerically, dR/dθ = −0.0264101v
2
0
). Atthispoint,
increasingtheanglewill decrease therange.
Showthat if
π
2
< θ < π, then thedistancealong thex-axis between θ and thepoint wherethetangent line
intersectsthex-axisisequal to|tanθ| (Figure4).
Further Insights and Challenges
53. Usethelimitdeﬁnitionof thederivativeandtheadditionlawfor thecosinefunctiontoprovethat(cosx)

= −sinx.
solution Let f (x) = cosx. Then
f

(x) = lim
h→0
cos(x +h) −cosx
h
= lim
h→0
cosx cosh −sinx sinh −cosx
h
= lim
h→0
_
(−sinx)
sinh
h
+(cosx)
cosh −1
h
_
= (−sinx) · 1+(cosx) · 0= −sinx.
Usetheadditionformulafor thetangent
tan(x +h) =
tanx +tanh
1+tanx tanh
tocompute(tanx)

directlyasalimit of thedifferencequotients. Youwill alsoneedtoshowthat lim
h→0
tanh
h
= 1.
55. Verifythefollowingidentityanduseit togiveanother proof of theformula(sinx)

= cosx.
sin(x +h) −sinx = 2cos
_
x +
1
2
h
_
sin
_
1
2
h
_
Hint: Usetheadditionformulatoprovethat sin(a +b) −sin(a −b) = 2cosa sinb.
solution Recall that
sin(a +b) = sina cosb +cosa sinb
and
sin(a −b) = sina cosb −cosa sinb.
Subtractingthesecondidentityfromtheﬁrst yields
sin(a +b) −sin(a −b) = 2cosa sinb.
If wenowset a = x +
h
2
andb =
h
2
, thenthepreviousequationbecomes
sin(x +h) −sinx = 2cos
_
x +
h
2
_
sin
_
h
2
_
.
Finally, weusethelimit deﬁnitionof thederivativeof sinx toobtain
d
dx
sinx = lim
h→0
sin(x +h) −sinx
h
= lim
h→0
2cos
_
x +
h
2
_
sin
_
h
2
_
h
= lim
h→0
cos
_
x +
h
2
_
· lim
h→0
sin
_
h
2
_
_
h
2
_ = cosx · 1= cosx.
Inother words,
d
dx
(sinx) = cosx.
Showthat anonzeropolynomial functiony = f (x) cannot satisfy theequationy

= −y. Usethistoprove
that neither sinx nor cosx is apolynomial. Canyouthink of another way to reachthis conclusionby considering
limitsasx → ∞?
57. Let f (x) = x sinx andg(x) = x cosx.
(a) Showthat f

(x) = g(x) +sinx andg

(x) = −f (x) +cosx.
(b) Verifythat f

(x) = −f (x) +2cosx and
g

(x) = −g(x) −2sinx.
(c) Byfurtherexperimentation, trytoﬁndformulasforall higherderivativesof f andg. Hint: Thekthderivativedepends
onwhether k = 4n, 4n +1, 4n +2, or 4n +3.
solution Let f (x) = x sinx andg(x) = x cosx.
(a) Weexamineﬁrst derivatives: f

(x) = x cosx +(sinx) · 1= g(x) +sinx andg

(x) = (x)(−sinx) +(cosx) · 1=
−f (x) +cosx; i.e., f

(x) = g(x) +sinx andg

(x) = −f (x) +cosx.
June 8, 2011 LTSV SSM Second Pass
138 C HA P T E R 3 DIFFERENTIATION
(b) Now look at second derivatives: f

(x) = g

(x) + cosx = −f (x) + 2cosx and g

(x) = −f

(x) − sinx =
−g(x) −2sinx; i.e., f

(x) = −f (x) +2cosx andg

(x) = −g(x) −2sinx.
(c) • Thethirdderivatives aref

(x) = −f

(x) − 2sinx = −g(x) − 3sinx andg

(x) = −g

(x) − 2cosx =
f (x) −3cosx; i.e., f

(x) = −g(x) −3sinx andg

(x) = f (x) −3cosx.
• The fourth derivatives are f
(4)
(x) = −g

(x) − 3cosx = f (x) − 4cosx and g
(4)
(x) = f

(x) + 3sinx =
g(x) +4sinx; i.e., f
(4)
= f (x) −4cosx andg
(4)
(x) = g(x) +4sinx.
• Wecannowseethepatternfor thederivatives, whicharesummarizedinthefollowingtable. Heren = 0, 1, 2, . . .
k 4n 4n +1 4n +2 4n +3
f
(k)
(x) f (x) −k cosx g(x) +k sinx −f (x) +k cosx −g(x) −k sinx
g
(k)
(x) g(x) +k sinx −f (x) +k cosx −g(x) −k sinx f (x) −k cosx
Figure5showsthegeometrybehindthederivativeformula(sinθ)

= cosθ. SegmentsBAandBD areparallel
tothex- andy-axes. Let sinθ = sin(θ +h) −sinθ. Verifythefollowingstatements.
(a) sinθ = BC
(b)

BDA = θ Hint: OA ⊥ AD.
(c) BD = (cosθ)AD
Nowexplainthefollowingintuitiveargument: If h issmall, thenBC ≈ BD andAD ≈ h, sosinθ ≈ (cosθ)h and
(sinθ)

= cosθ.
3.7 The Chain Rule
Preliminary Questions
1. Identifytheoutsideandinsidefunctionsfor eachof thesecompositefunctions.
(a) y =
_
4x +9x
2
(b) y = tan(x
2
+1)
(c) y = sec
5
x (d) y = (1+x
12
)
4
solution
(a) Theouter functionis

x, andtheinner functionis4x +9x
2
.
(b) Theouter functionistanx, andtheinner functionisx
2
+1.
(c) Theouter functionisx
5
, andtheinner functionissecx.
(d) Theouter functionisx
4
, andtheinner functionis1+x
12
.
2. Whichof thefollowingcanbedifferentiatedeasilywithout usingtheChainRule?
(a) y = tan(7x
2
+2) (b) y =
x
x +1
(c) y =

x · secx (d) y =

x cosx
(e) y = x sec

x (f) y = tan(4x)
solution Thefunction
x
x+1
canbedifferentiatedusingtheQuotient Rule, andthefunction

x · secx canbediffer-
entiatedusingtheProductRule. Thefunctionstan(7x
2
+2),

x cosx, andtan(4x) requiretheChainRule. x sec

x can
bepartiallyevaluatedusingtheProductRule, butthentheChainruleisneededtodifferentiateoneof thefactors, sec

x.
3. Whichisthederivativeof f (5x)?
(a) 5f

(x) (b) 5f

(5x) (c) f

(5x)
solution Thecorrect answer is(b): 5f

(5x).
4. Supposethatf

(4) = g(4) = g

(4) = 1. DowehaveenoughinformationtocomputeF

(4), whereF(x) = f (g(x))?
If not, what ismissing?
solution If F(x) = f (g(x)), thenF

(x) = f

(g(x))g

(x) andF

(4) = f

(g(4))g

(4). Thus, wedonothaveenough
informationtocomputeF

(4). Wearemissingthevalueof f

(1).
Exercises
In Exercises 1–4, ﬁll in a table of the following type:
f (g(x)) f

(u) f

(g(x)) g

(x) (f ◦ g)

1. f (u) = u
3/2
, g(x) = x
4
+1
solution
f (g(x)) f

(u) f

(g(x)) g

(x) (f ◦ g)

(x
4
+1)
3/2 3
2
u
1/2 3
2
(x
4
+1)
1/2
4x
3
6x
3
(x
4
+1)
1/2
June 8, 2011 LTSV SSM Second Pass
S E C T I ON 3.7 The Chain Rule 139
f (u) = u
3
, g(x) = 3x +5
3. f (u) = tanu, g(x) = x
4
solution
f (g(x)) f

(u) f

(g(x)) g

(x) (f ◦ g)

tan(x
4
) sec
2
u sec
2
(x
4
) 4x
3
4x
3
sec
2
(x
4
)
f (u) = u
4
+u, g(x) = cosx
In Exercises 5 and 6, write the function as a composite f (g(x)) and compute the derivative using the Chain Rule.
5. y = (x +sinx)
4
solution Let f (x) = x
4
, g(x) = x +sinx, andy = f (g(x)) = (x +sinx)
4
. Then
dy
dx
= f

(g(x))g

(x) = 4(x +sinx)
3
(1+cosx).
y = cos(x
3
)
7. Calculate
d
dx
cosu for thefollowingchoicesof u(x):
(a) u = 9−x
2
(b) u = x
−1
(c) u = tanx
solution
(a) cos(u(x)) = cos(9−x
2
).
d
dx
cos(u(x)) = −sin(u(x))u

(x) = −sin(9−x
2
)(−2x) = 2x sin(9−x
2
).
(b) cos(u(x)) = cos(x
−1
).
d
dx
cos(u(x)) = −sin(u(x))u

(x) = −sin(x
−1
)
_

1
x
2
_
=
sin(x
−1
)
x
2
.
(c) cos(u(x)) = cos(tanx).
d
dx
cos(u(x)) = −sin(u(x))u

(x) = −sin(tanx)(sec
2
x) = −sec
2
x sin(tanx).
Calculate
d
dx
f (x
2
+1) for thefollowingchoicesof f (u):
(a) f (u) = sinu (b) f (u) = 3u
3/2
(c) f (u) = u
2
−u
9. Compute
df
dx
if
df
du
= 2and
du
dx
= 6.
solution Assumingf isafunctionof u, whichisinturnafunctionof x,
df
dx
=
df
du
·
du
dx
= 2(6) = 12.
Compute
df
dx
¸
¸
¸
x=2
if f (u) = u
2
, u(2) = −5, andu

(2) = −5.
In Exercises 11–22, use the General Power Rule or the Shifting and Scaling Rule to compute the derivative.
11. y = (x
4
+5)
3
solution UsingtheGeneral Power Rule,
d
dx
(x
4
+5)
3
= 3(x
4
+5)
2
d
dx
(x
4
+5) = 3(x
4
+5)
2
(4x
3
) = 12x
3
(x
4
+5)
2
.
y = (8x
4
+5)
3
13. y =

7x −3
solution UsingtheShiftingandScalingRule
d
dx

7x −3=
d
dx
(7x −3)
1/2
=
1
2
(7x −3)
−1/2
(7) =
7
2

7x −3
.
y = (4−2x −3x
2
)
5
15. y = (x
2
+9x)
−2
solution UsingtheGeneral Power Rule,
d
dx
(x
2
+9x)
−2
= −2(x
2
+9x)
−3
d
dx
(x
2
+9x) = −2(x
2
+9x)
−3
(2x +9).
y = (x
3
+3x +9)
−4/3
17. y = cos
4
θ
solution UsingtheGeneral Power Rule,
d

cos
4
θ = 4cos
3
θ
d

cosθ = −4cos
3
θ sinθ.
June 8, 2011 LTSV SSM Second Pass
140 C HA P T E R 3 DIFFERENTIATION
y = cos(9θ +41)
19. y = (2cosθ +5sinθ)
9
solution UsingtheGeneral Power Rule,
d

(2cosθ +5sinθ)
9
= 9(2cosθ +5sinθ)
8
d

(2cosθ +5sinθ) = 9(2cosθ +5sinθ)
8
(5cosθ −2sinθ).
y =

9+x +sinx
21. y = sin
__
x
2
+2x +9
_
solution Usingthegeneral power rule,
d
dx
sin
__
x
2
+2x +9
_
= cos
__
x
2
+2x +9
_
·
1
2
(x
2
+2x +9)
−1/2
· (2x +2)
= (x +1)(x
2
+2x +9)
−1/2
cos
__
x
2
+2x +9
_
y = tan(4−3x) sec(3−4x)
In Exercises 23–26, compute the derivative of f ◦ g.
23. f (u) = sinu, g(x) = 2x +1
solution Let h(x) = f (g(x)) = sin(2x + 1). Then, applyingtheshiftingandscalingrule, h

(x) = 2cos(2x + 1).
Alternately,
d
dx
f (g(x)) = f

(g(x))g

(x) = cos(2x +1) · 2= 2cos(2x +1).
f (u) = 2u +1, g(x) = sinx
25. f (u) = u +u
−1
, g(x) = tanx
solution Let h(x) = f (g(x)) = tanx +cotx. Thenh

(x) = sec
2
x −csc
2
x. Alternatively,
d
dx
f (g(x)) = f

(g(x))g

(x) = (1−cot
2
x) sec
2
x = sec
2
x −csc
2
x
f (u) =
u
u −1
, g(x) = cscx
In Exercises 27 and 28, ﬁnd the derivatives of f (g(x)) and g(f (x)).
27. f (u) = cosu, u = g(x) = x
2
+1
solution
d
dx
f (g(x)) = f

(g(x))g

(x) = −sin(x
2
+1)(2x) = −2x sin(x
2
+1).
d
dx
g(f (x)) = g

(f (x))f

(x) = 2(cosx)(−sinx) = −2sinx cosx.
f (u) = u
3
, u = g(x) =
1
x +1
In Exercises 29–42, use the Chain Rule to ﬁnd the derivative.
29. y = sin(x
2
)
solution Let y = sin
_
x
2
_
. Theny

= cos
_
x
2
_
· 2x = 2x cos
_
x
2
_
.
y = sin
2
x
31. y =
_
t
2
+9
solution Let y =
_
t
2
+9= (t
2
+9)
1/2
. Then
y

=
1
2
(t
2
+9)
−1/2
(2t ) =
t
_
t
2
+9
.
y = (t
2
+3t +1)
−5/2
33. y = (x
4
−x
3
−1)
2/3
solution Let y =
_
x
4
−x
3
−1
_
2/3
. Then
y

=
2
3
_
x
4
−x
3
−1
_
−1/3
_
4x
3
−3x
2
_
.
y = (

x +1−1)
3/2
35. y =
_
x +1
x −1
_
4
solution Let y =
_
x +1
x −1
_
4
. Then
y

= 4
_
x +1
x −1
_
3
·
(x −1) · 1−(x +1) · 1
(x −1)
2
= −
8(x +1)
3
(x −1)
5
=
8(1+x)
3
(1−x)
5
.
June 8, 2011 LTSV SSM Second Pass
S E C T I ON 3.7 The Chain Rule 141
y = cos
3
(12θ)
37. y = sec
1
x
solution Let f (x) = sec
_
x
−1
_
. Then
f

(x) = sec
_
x
−1
_
tan
_
x
−1
_
·
_
−x
−2
_
= −
sec(1/x) tan(1/x)
x
2
.
y = tan(θ
2
−4θ)
39. y = tan(θ +cosθ)
solution Let y = tan(θ +cosθ). Then
y

= sec
2
(θ +cosθ) · (1−sinθ) = (1−sinθ) sec
2
(θ +cosθ) .
y =
_
cot
9
θ +1
41. y = csc(9−2θ
2
)
solution Wehave
y

= −csc(9−2θ
2
) cot(9−2θ
2
) · (−4θ) = 4θ csc(9−2θ
2
) cot(9−2θ
2
)
y = cot(

θ −1)
In Exercises 43–72, ﬁnd the derivative using the appropriate rule or combination of rules.
43. y = tan(x
2
+4x)
solution Let y = tan(x
2
+4x). Bythechainrule,
y

= sec
2
(x
2
+4x) · (2x +4) = (2x +4) sec
2
(x
2
+4x).
y = sin(x
2
+4x)
45. y = x cos(1−3x)
solution Let y = x cos(1−3x). Applyingtheproduct ruleandthenthescalingandshiftingrule,
y

= x (−sin(1−3x)) · (−3) +cos(1−3x) · 1= 3x sin(1−3x) +cos(1−3x) .
y = sin(x
2
) cos(x
2
)
47. y = (4t +9)
1/2
solution Let y = (4t +9)
1/2
. Bytheshiftingandscalingrule,
dy
dt
= 4
_
1
2
_
(4t +9)
−1/2
= 2(4t +9)
−1/2
.
y = (z +1)
4
(2z −1)
3
49. y = (x
3
+cosx)
−4
solution Let y = (x
3
+cosx)
−4
. Bythegeneral power rule,
y

= −4(x
3
+cosx)
−5
(3x
2
−sinx) = 4(sinx −3x
2
)(x
3
+cosx)
−5
.
y = sin(cos(sinx))
51. y =

sinx cosx
solution Westart byusingatrigidentitytorewrite
y =

sinx cosx =
_
1
2
sin2x =
1

2
(sin2x)
1/2
.
Then, after twoapplicationsof thechainrule,
y

=
1

2
·
1
2
(sin2x)
−1/2
· cos2x · 2=
cos2x

2sin2x
.
y = (9−(5−2x
4
)
7
)
3
53. y = (cos6x +sinx
2
)
1/2
solution Let y = (cos6x + sin(x
2
))
1/2
. Applyingthegeneral power rulefollowedby boththescalingandshifting
ruleandthechainrule,
y

=
1
2
_
cos6x +sin(x
2
)
_
−1/2
_
−sin6x · 6+cos(x
2
) · 2x
_
=
x cos(x
2
) −3sin6x
_
cos6x +sin(x
2
)
.
June 8, 2011 LTSV SSM Second Pass
142 C HA P T E R 3 DIFFERENTIATION
y =
(x +1)
1/2
x +2
55. y = tan
3
x +tan(x
3
)
solution Let y = tan
3
x + tan(x
3
) = (tanx)
3
+ tan(x
3
). Applyingthegeneral power ruletotheﬁrst termandthe
chainruletothesecondterm,
y

= 3(tanx)
2
sec
2
x +sec
2
(x
3
) · 3x
2
= 3
_
x
2
sec
2
(x
3
) +sec
2
x tan
2
x
_
.
y =

4−3cosx 57. y =
_
z +1
z −1
solution Let y =
_
z +1
z −1
_
1/2
. Applyingthegeneral power rulefollowedbythequotient rule,
dy
dz
=
1
2
_
z +1
z −1
_
−1/2
·
(z −1) · 1−(z +1) · 1
(z −1)
2
=
−1

z +1(z −1)
3/2
.
y = (cos
3
x +3cosx +7)
9 59. y =
cos(1+x)
1+cosx
solution Let
y =
cos(1+x)
1+cosx
.
Then, applyingthequotient ruleandtheshiftingandscalingrule,
dy
dx
=
−(1+cosx) sin(1+x) +cos(1+x) sinx
(1+cosx)
2
=
cos(1+x) sinx −cosx sin(1+x) −sin(1+x)
(1+cosx)
2
=
sin(−1) −sin(1+x)
(1+cosx)
2
.
Thelast linefollowsfromtheidentity
sin(A −B) = sinAcosB −cosAsinB
withA = x andB = 1+x.
y = sec(
_
t
2
−9)
61. y = cot
7
(x
5
)
solution Let y = cot
7
_
x
5
_
. Applyingthegeneral power rulefollowedbythechainrule,
dy
dx
= 7cot
6
_
x
5
_
·
_
−csc
2
_
x
5
__
· 5x
4
= −35x
4
cot
6
_
x
5
_
csc
2
_
x
5
_
.
y =
cos(1/x)
1+x
2
63. y =
_
1+cot
5
(x
4
+1)
_
9
solution Lety =
_
1+cot
5
_
x
4
+1
__
9
. Applyingthegeneral power rule, thechainrule, andthegeneral power rule
insuccession,
dy
dx
= 9
_
1+cot
5
_
x
4
+1
__
8
· 5cot
4
_
x
4
+1
_
·
_
−csc
2
_
x
4
+1
__
· 4x
3
= −180x
3
cot
4
_
x
4
+1
_
csc
2
_
x
4
+1
_ _
1+cot
5
_
x
4
+1
__
8
.
y =

cos2x +sin4x
65. y = (1−csc
2
(1−x
3
))
6
solution Usingthechainrulemultipletimes, wehave
d
dx
(1−csc
2
(1−x
3
))
6
= 6(1−csc
2
(1−x
3
))
5
d
dx
(1−csc
2
(1−x
3
))
= 6(1−csc
2
(1−x
3
))
5
(−2csc(1−x
3
))(−csc(1−x
3
) cot(1−x
3
))(−3x
2
)
= −36x
2
csc
2
(1−x
3
) cot(1−x
3
)(1−csc
2
(1−x
3
))
5
y = sin(

sinθ +1)
67. y =
_
x +
1
x
_
−1/2
solution Applyingthechainrulegives
d
dx
_
x +
1
x
_
−1/2
= −
1
2
_
x +
1
x
_
−3/2
(1−x
−2
) =
1
2
(x
−2
−1)(x +x
−1
)
−3/2
June 8, 2011 LTSV SSM Second Pass
S E C T I ON 3.7 The Chain Rule 143
y = sec(1+(4+x)
−3/2
)
69. y =
_
1+
_
1+

x
solution Let y =
_
1+
_
1+x
1/2
_
1/2
_
1/2
. Applyingthegeneral power ruletwice,
dy
dx
=
1
2
_
1+
_
1+x
1/2
_
1/2
_
−1/2
·
1
2
_
1+x
1/2
_
−1/2
·
1
2
x
−1/2
=
1
8

x
_
1+

x
_
1+
_
1+

x
.
y =
_

x +1+1
71. y = (kx +b)
−1/3
; k andb anyconstants
solution Let y = (kx +b)
−1/3
, whereb andk areconstants. Bythescalingandshiftingrule,
y

= −
1
3
(kx +b)
−4/3
· k = −
k
3
(kx +b)
−4/3
.
y =
1
_
kt
4
+b
; k, b constants, not bothzero
In Exercises 73–76, compute the higher derivative.
73.
d
2
dx
2
sin(x
2
)
solution Let f (x) = sin
_
x
2
_
. Then, bythechainrule, f

(x) = 2x cos
_
x
2
_
and, bytheproduct ruleandthechain
rule,
f

(x) = 2x
_
−sin
_
x
2
_
· 2x
_
+2cos
_
x
2
_
= 2cos
_
x
2
_
−4x
2
sin
_
x
2
_
.
d
2
dx
2
(x
2
+9)
5
75.
d
3
dx
3
(9−x)
8
solution Let f (x) = (9−x)
8
. Then, byrepeateduseof thescalingandshiftingrule,
f

(x) = 8(9−x)
7
· (−1) = −8(9−x)
7
f

(x) = −56(9−x)
6
· (−1) = 56(9−x)
6
,
f

(x) = 336(9−x)
5
· (−1) = −336(9−x)
5
.
d
3
dx
3
sin(2x)
77. The average molecular velocity v of a gas in a certain container is given by v = 29

T m/s, where T is the
temperatureinkelvins. Thetemperatureisrelatedtothepressure(inatmospheres) byT = 200P. Find
dv
dP
¸
¸
¸
¸
P=1.5
.
solution First notethat whenP = 1.5atmospheres, T = 200(1.5) = 300K. Thus,
dv
dP
¸
¸
¸
¸
P=1.5
=
dv
dT
¸
¸
¸
¸
T =300
·
dT
dP
¸
¸
¸
¸
P=1.5
=
29
2

300
· 200=
290

3
3
m
s· atmospheres
.
Alternately, substitutingT = 200P intotheequationfor v givesv = 290

2P. Therefore,
dv
dP
=
290

2
2

P
=
290

2P
,
so
dv
dP
¸
¸
¸
¸
P=1.5
=
290

3
=
290

3
3
m
s· atmospheres
.
Thepower P inacircuit is P = Ri
2
, whereR is theresistanceandi is thecurrent. FinddP/dt at t =
1
3
if
R = 1000 andi variesaccordingtoi = sin(4πt ) (timeinseconds).
79. Anexpandingspherehas radius r = 0.4t cmat timet (inseconds). Let V bethesphere’s volume. FinddV/dt
when(a) r = 3and(b) t = 3.
solution Let r = 0.4t , wheret isinseconds(s) andr isincentimeters(cm). WithV =
4
3
πr
3
, wehave
dV
dr
= 4πr
2
.
Thus
dV
dt
=
dV
dr
dr
dt
= 4πr
2
· (0.4) = 1.6πr
2
.
(a) Whenr = 3,
dV
dt
= 1.6π(3)
2
≈ 45.24cm/s.
(b) Whent = 3, wehaver = 1.2. Hence
dV
dt
= 1.6π(1.2)
2
≈ 7.24cm/s.
June 8, 2011 LTSV SSM Second Pass
144 C HA P T E R 3 DIFFERENTIATION
A 2005study by theFisheriesResearchServicesinAberdeen, Scotland, suggeststhat theaveragelengthof the
speciesClupea harengus (Atlanticherring) asafunctionof aget (inyears) canbemodeledby
L(t ) = 32
_
1−(1+0.37t +0.068t
2
+0.0085t
3
+0.0009t
4
)
−1
_
for 0≤ t ≤ 13. SeeFigure2.
(a) Howfast istheaveragelengthchangingat aget = 6years?
(b) At what ageistheaveragelengthchangingat arateof 5cm/yr?
81. According to a1999 study by Starkey and Scarnecchia, theaverageweight (in kilograms) at aget (in years) of
channel catﬁshintheLower YellowstoneRiver (Figure3) isapproximatedbythefunction:
W(t ) = (0.14+0.115t −0.002t
2
+0.000023t
3
)
3.4
Findtherateat whichaverageweight ischangingat aget = 10.
5 10 15 20
1
2
3
4
5
6
7
8
t (year)
W (kg)
Lower YellowstoneRiver
FIGURE 3 Averageweight of channel catﬁshat aget
solution Let W(t ) = (0.14+0.115t −0.002t
2
+0.000023t
3
)
3.4
. Then
W

(t ) = 3.4(0.14+0.115t −0.002t
2
+0.000023t
3
)
2.4
(0.115−0.004t +0.000069t
2
)
At aget = 10,
W

(10) ≈ 0.36kg/yr.
CalculateM

(0) intermsof theconstantsa, b, k, andm, where
M(t ) =
_
a +(b −a)
_
1+kmt +
1
2
(kmt )
2
__
1/m
83. WithnotationasinExample7, calculate
(a)
d

sinθ
¸
¸
¸
¸
θ=60

(b)
d

_
θ +tanθ
_
¸
¸
¸
¸
θ=45

solution
(a)
d

sinθ
¸
¸
¸
θ=60

=
d

sin
_
π
180
θ
_ ¸
¸
¸
θ=60

=
_
π
180
_
cos
_
π
180
(60)
_
=
π
180
1
2
=
π
360
.
(b)
d

_
θ +tanθ
_
¸
¸
¸
θ=45

=
d

_
θ +tan
_
π
180
θ
__ ¸
¸
¸
θ=45

= 1+
π
180
sec
2
_
π
4
_
= 1+
π
90
.
Assumethat
f (0) = 2, f

(0) = 3, h(0) = −1, h

(0) = 7
Calculatethederivativesof thefollowingfunctionsat x = 0:
(a) (f (x))
3
(b) f (7x) (c) f (4x)h(5x)
85. Computethederivativeof h(sinx) at x =
π
6
, assumingthat h

(0.5) = 10.
solution Let u = sinx andsupposethat h

(0.5) = 10. Then
d
dx
(h(u)) =
dh
du
du
dx
=
dh
du
cosx.
Whenx =
π
6
, wehaveu = 0.5. Accordingly, thederivativeof h(sinx) at x =
π
6
is10cos
_
π
6
_
= 5

3.
Let F(x) = f (g(x)), wherethegraphs of f and g areshown in Figure4. Estimateg

(2) and f

(g(2)) and
computeF

(2).
In Exercises 87–90, use the table of values to calculate the derivative of the function at the given point.
x 1 4 6
f (x) 4 0 6
f

(x) 5 7 4
g(x) 4 1 6
g

(x) 5
1
2
3
87. f (g(x)), x = 6
solution
d
dx
f (g(x))
¸
¸
¸
¸
x=6
= f

(g(6))g

(6) = f

(6)g

(6) = 4×3= 12.
sin(πg(x)), x = 4
89. g(

x), x = 16
solution
d
dx
g(

x)
¸
¸
¸
¸
x=16
= g

(4)
_
1
2
_
(1/

16) =
_
1
2
__
1
2
__
1
4
_
=
1
16
.
f (2x +g(x)), x = 1
June 8, 2011 LTSV SSM Second Pass
S E C T I ON 3.7 The Chain Rule 145
91. Theprice(indollars) of acomputer componentisP = 2C −18C
−1
, whereC isthemanufacturer’scosttoproduce
it. Assumethat cost at timet (inyears) isC = 9+ 3t
−1
. Determinetherateof changeof pricewithrespect totimeat
t = 3.
solution
dC
dt
= −3t
−2
. C(3) = 10andC

(3) = −
1
3
, sowecompute:
dP
dt
¸
¸
¸
¸
t =3
= 2C

(3) +
18
(C(3))
2
C

(3) = −
2
3
+
18
100
_

1
3
_
= −0.727
dollars
year
.
Plotthe“astroid”y = (4−x
2/3
)
3/2
for0≤ x ≤ 8. Showthatthepartof everytangentlineintheﬁrstquadrant
hasaconstant length8.
93. AccordingtotheU.S. standardatmospheric model, developedby theNational Oceanic andAtmosphericAdmin-
istrationfor useinaircraft androcket design, atmospherictemperatureT (indegreesCelsius), pressureP (kPa= 1000
pascals), andaltitudeh (inmeters) arerelatedbytheseformulas(validinthetroposphereh ≤ 11,000):
T = 15.04−0.000649h, P = 101.29+
_
T +273.1
288.08
_
5.256
UsetheChainRuletocalculatedP/dh. ThenestimatethechangeinP (inpascals, Pa) per additional meter of altitude
whenh = 3000.
solution
dP
dT
= 5.256
_
T +273.1
288.08
_
4.256
_
1
288.08
_
= 6.21519×10
−13
(273.1+T )
4.256
and
dT
dh
= −0.000649

C/m.
dP
dh
=
dP
dT
dT
dh
, so
dP
dh
=
_
6.21519×10
−13
(273.1+T )
4.256
_
(−0.000649) = −4.03366×10
−16
(288.14−0.000649h)
4.256
.
Whenh = 3000,
dP
dh
= −4.03366×10
−16
(286.193)
4.256
= −1.15×10
−5
kPa/m;
therefore, for eachadditional meter of altitude,
P ≈ −1.15×10
−5
kPa= −1.15×10
−2
Pa.
ClimatescientistsusetheStefan-BoltzmannLawR = σT
4
toestimatethechangeintheearth’saveragetemper-
atureT (inkelvins) causedbyachangeintheradiationR (injoulespersquaremeterpersecond) thattheearthreceives
fromthesun. Hereσ = 5.67×10
−8
J s
−1
m
−2
K
−4
. CalculatedR/dt , assumingthatT = 283and
dT
dt
= 0.05K/yr.
What aretheunitsof thederivative?
95. Inthesettingof Exercise94, calculatetheyearlyrateof changeof T if T = 283K andR increasesat arateof 0.5
J s
−1
m
−2
per year.
solution BytheChainRule,
dR
dt
=
dR
dT
·
dT
dt
= 4σT
3
dT
dt
.
AssumingT = 283K and
dR
dt
= 0.5J s
−1
m
−2
per year, it followsthat author:
0.5= 4σ(283)
3
dT
dt

dT
dt
=
0.5
4σ(283)
3
≈ 0.0973 kelvins/yr
Useacomputer algebrasystemtocomputef
(k)
(x) for k = 1, 2, 3for thefollowingfunctions:
(a) f (x) = cot(x
2
) (b) f (x) =
_
x
3
+1
97. UsetheChainRuletoexpressthesecondderivativeof f ◦ g intermsof theﬁrstandsecondderivativesof f andg.
solution Let h(x) = f (g(x)). Then
h

(x) = f

(g(x))g

(x)
and
h

(x) = f

(g(x))g

(x) +g

(x)f

(g(x))g

(x) = f

(g(x))g

(x) +f

(g(x))
_
g

(x)
_
2
.
Computethesecondderivativeof sin(g(x)) at x = 2, assumingthat g(2) =
π
4
, g

(2) = 5, andg

(2) = 3.
Further Insights and Challenges
99. Showthat if f , g, andh aredifferentiable, then
[f (g(h(x)))]

= f

(g(h(x)))g

(h(x))h

(x)
solution Let f , g, andh bedifferentiable. Let u = h(x), v = g(u), andw = f (v). Then
dw
dx
=
df
dv
dv
dx
=
df
dv
dg
du
du
dx
= f

(g(h(x))g

(h(x))h

(x)
June 8, 2011 LTSV SSM Second Pass
146 C HA P T E R 3 DIFFERENTIATION
Showthat differentiationreverses parity: If f is even, thenf

is odd, andif f is odd, thenf

is even. Hint:
Differentiatef (−x).
101. (a) Sketchagraphof anyevenfunctionf (x) andexplaingraphicallywhyf

(x) isodd.
(b) Supposethat f

(x) iseven. Isf (x) necessarilyodd? Hint: Checkwhether thisistruefor linear functions.
solution
(a) Thegraphof anevenfunctionissymmetricwithrespect tothey-axis. Accordingly, itsimageintheleft half-planeis
amirror reﬂectionof that intheright half-planethroughthey-axis. If at x = a ≥ 0, theslopeof f existsandisequal to
m, thenby reﬂectionitsslopeat x = −a ≤ 0is−m. That is, f

(−a) = −f

(a). Note: Thismeansthat if f

(0) exists,
thenit equals0.
y
x
−2 −1 1 2
1
2
3
4
(b) Supposethat f

iseven. Thenf isnot necessarilyodd. Let f (x) = 4x +7. Thenf

(x) = 4, anevenfunction. But
f isnot odd. For example, f (2) = 15, f (−2) = −1, but f (−2) = −f (2).
Power Rulefor Fractional Exponents Let f (u) = u
q
andg(x) = x
p/q
. Assumethat g(x) isdifferentiable.
(a) Showthat f (g(x)) = x
p
(recall thelawsof exponents).
(b) ApplytheChainRuleandthePower Rulefor whole-number exponentstoshowthat f

(g(x)) g

(x) = px
p−1
.
(c) ThenderivethePower Rulefor x
p/q
.
103. Provethat for all wholenumbersn ≥ 1,
d
n
dx
n
sinx = sin
_
x +

2
_
Hint: Usetheidentitycosx = sin
_
x +
π
2
_
.
solution Wewill proceedbyinductiononn. For n = 1, weﬁnd
d
dx
sinx = cosx = sin
_
x +
π
2
_
,
asrequired. Now, supposethat for somepositiveinteger k,
d
k
dx
k
sinx = sin
_
x +

2
_
.
Then
d
k+1
dx
k+1
sinx =
d
dx
sin
_
x +

2
_
= cos
_
x +

2
_
= sin
_
x +
(k +1)π
2
_
.
ADiscontinuousDerivative Usethelimit deﬁnitiontoshowthat g

(0) existsbut g

(0) = lim
x→0
g

(x), where
g(x) =

x
2
sin
1
x
x = 0
0 x = 0
105. ChainRule ThisexerciseprovestheChainRulewithoutthespecial assumptionmadeinthetext. For anynumber
b, deﬁneanewfunction
F(u) =
f (u) −f (b)
u −b
for all u = b
(a) Showthat if wedeﬁneF(b) = f

(b), thenF(u) iscontinuousat u = b.
(b) Takeb = g(a). Showthat if x = a, thenfor all u,
f (u) −f (g(a))
x −a
= F(u)
u −g(a)
x −a
2
Notethat bothsidesarezeroif u = g(a).
(c) Substituteu = g(x) inEq. (2) toobtain
f (g(x)) −f (g(a))
x −a
= F(g(x))
g(x) −g(a)
x −a
DerivetheChainRulebycomputingthelimit of bothsidesasx → a.
solution For anydifferentiablefunctionf andanynumber b, deﬁne
F(u) =
f (u) −f (b)
u −b
for all u = b.
June 8, 2011 LTSV SSM Second Pass
S E C T I ON 3.8 Implicit Differentiation 147
(a) DeﬁneF(b) = f

(b). Then
lim
u→b
F(u) = lim
u→b
f (u) −f (b)
u −b
= f

(b) = F(b),
i.e., lim
u→b
F(u) = F(b). Therefore, F iscontinuousat u = b.
(b) Let g beadifferentiablefunctionandtakeb = g(a). Let x beanumber distinct froma. If wesubstituteu = g(a)
intoEq. (2), bothsidesevaluateto0, soequalityissatisﬁed. Ontheother hand, if u = g(a), then
f (u) −f (g(a))
x −a
=
f (u) −f (g(a))
u −g(a)
u −g(a)
x −a
=
f (u) −f (b)
u −b
u −g(a)
x −a
= F(u)
u −g(a)
x −a
.
(c) Hencefor all u, wehave
f (u) −f (g(a))
x −a
= F(u)
u −g(a)
x −a
.
(d) Substitutingu = g(x) inEq. (2), wehave
f (g(x)) −f (g(a))
x −a
= F(g(x))
g(x) −g(a)
x −a
.
Lettingx → a gives
lim
x→a
f (g(x)) −f (g(a))
x −a
= lim
x→a
_
F(g(x))
g(x) −g(a)
x −a
_
= F(g(a))g

(a) = F(b)g

(a) = f

(b)g

(a)
= f

(g(a))g

(a)
Therefore(f ◦ g)

(a) = f

(g(a))g

(a), whichistheChainRule.
3.8 Implicit Differentiation
Preliminary Questions
1. Whichdifferentiationruleisusedtoshow
d
dx
siny = cosy
dy
dx
?
solution Thechainruleisusedtoshowthat
d
dx
siny = cosy
dy
dx
.
2. Oneof (a)–(c) isincorrect. Findandcorrect themistake.
(a)
d
dy
sin(y
2
) = 2y cos(y
2
) (b)
d
dx
sin(x
2
) = 2x cos(x
2
) (c)
d
dx
sin(y
2
) = 2y cos(y
2
)
solution
(a) Thisiscorrect. Notethat thedifferentiationiswithrespect tothevariabley.
(b) Thisiscorrect. Notethat thedifferentiationiswithrespect tothevariablex.
(c) Thisisincorrect. Becausethedifferentiationiswithrespect tothevariablex, thechainruleisneededtoobtain
d
dx
sin(y
2
) = 2y cos(y
2
)
dy
dx
.
3. Onanexam, J asonwasaskedtodifferentiatetheequation
x
2
+2xy +y
3
= 7
FindtheerrorsinJ ason’sanswer: 2x +2xy

+3y
2
= 0
solution Therearetwo mistakes inJ ason’s answer. First, J asonshouldhaveappliedtheproduct ruleto thesecond
termtoobtain
d
dx
(2xy) = 2x
dy
dx
+2y.
Second, heshouldhaveappliedthegeneral power ruletothethirdtermtoobtain
d
dx
y
3
= 3y
2
dy
dx
.
June 8, 2011 LTSV SSM Second Pass
148 C HA P T E R 3 DIFFERENTIATION
4. Whichof (a) or (b) isequal to
d
dx
(x sint )?
(a) (x cost )
dt
dx
(b) (x cost )
dt
dx
+sint
solution Usingtheproduct ruleandthechainruleweseethat
d
dx
(x sint ) = x cost
dt
dx
+sint,
sothecorrect answer is(b).
Exercises
1. Showthat if youdifferentiatebothsidesof x
2
+2y
3
= 6, theresult is2x +6y
2
dy
dx
= 0. Thensolvefor dy/dx and
evaluateit at thepoint (2, 1).
solution
d
dx
(x
2
+2y
3
) =
d
dx
6
2x +6y
2
dy
dx
= 0
2x +6y
2
dy
dx
= 0
6y
2
dy
dx
= −2x
dy
dx
=
−2x
6y
2
.
At (2, 1),
dy
dx
=
−4
6
= −
2
3
.
Showthat if youdifferentiatebothsidesof xy +4x +2y = 1, theresult is(x +2)
dy
dx
+y +4= 0. Thensolve
for dy/dx andevaluateit at thepoint (1, −1).
In Exercises 3–8, differentiate the expression with respect to x, assuming that y = f (x).
3. x
2
y
3
solution Assumingthat y dependsonx, then
d
dx
_
x
2
y
3
_
= x
2
· 3y
2
y

+y
3
· 2x = 3x
2
y
2
y

+2xy
3
.
x
3
y
2
5. (x
2
+y
2
)
3/2
solution Assumingthat y dependsonx, then
d
dx
_
_
x
2
+y
2
_
3/2
_
=
3
2
_
x
2
+y
2
_
1/2_
2x +2yy

_
= 3
_
x +yy

_
_
x
2
+y
2
.
tan(xy)
7.
y
y +1
solution Assumingthat y dependsonx, then
d
dx
y
y +1
=
(y +1)y

−yy

(y +1)
2
=
y

(y +1)
2
.
sin
y
x
In Exercises 9–26, calculate the derivative with respect to x.
9. 3y
3
+x
2
= 5
solution Let 3y
3
+x
2
= 5. Then9y
2
y

+2x = 0, andy

= −
2x
9y
2
.
y
4
−2y = 4x
3
+x
11. x
2
y +2x
3
y = x +y
solution Let x
2
y +2x
3
y = x +y. Then
x
2
y

+2xy +2x
3
y

+6x
2
y = 1+y

x
2
y

+2x
3
y

−y

= 1−2xy −6x
2
y
y

=
1−2xy −6x
2
y
x
2
+2x
3
−1
.
xy
2
+x
2
y
5
−x
3
= 3
13. x
3
R
5
= 1
solution Let x
3
R
5
= 1. Thenx
3
· 5R
4
R

+R
5
· 3x
2
= 0, andR

= −
3x
2
R
5
5x
3
R
4
= −
3R
5x
.
June 8, 2011 LTSV SSM Second Pass
S E C T I ON 3.8 Implicit Differentiation 149
x
4
+z
4
= 1
15.
y
x
+
x
y
= 2y
solution Let
y
x
+
x
y
= 2y.
Then
xy

−y
x
2
+
y −xy

y
2
= 2y

_
1
x

x
y
2
−2
_
y

=
y
x
2

1
y
y
2
−x
2
−2xy
2
xy
2
y

=
y
2
−x
2
x
2
y
y

=
y(y
2
−x
2
)
x(y
2
−x
2
−2xy
2
)
.

x +s =
1
x
+
1
s
17. y
−2/3
+x
3/2
= 1
solution Let y
−2/3
+x
3/2
= 1. Then

2
3
y
−5/3
y

+
3
2
x
1/2
= 0 or y

=
9
4
x
1/2
y
5/3
.
x
1/2
+y
2/3
= −4y
19. y +
1
y
= x
2
+x
solution Let y +
1
y
= x
2
+x. Then
y

1
y
2
y

= 2x +1 or y

=
2x +1
1−y
−2
=
(2x +1)y
2
y
2
−1
.
sin(xt ) = t
21. sin(x +y) = x +cosy
solution Let sin(x +y) = x +cosy. Then
(1+y

) cos(x +y) = 1−y

siny
cos(x +y) +y

cos(x +y) = 1−y

siny
(cos(x +y) +siny) y

= 1−cos(x +y)
y

=
1−cos(x +y)
cos(x +y) +siny
.
tan(x
2
y) = (x +y)
3
23. tan(x +y) = tanx +tany
solution Implicitlydifferentiatinggives
sec
2
(x +y) +y

sec
2
(x +y) = sec
2
x +y

sec
2
y
y

(sec
2
(x +y) −sec
2
y) = sec
2
x −sec
2
(x +y)
y

=
sec
2
x −sec
2
(x +y)
sec
2
(x +y) −sec
2
y
x siny −y cosx = 2
25. x +cos(3x −y) = xy
solution Differentiateimplicitlytoget
1−3sin(3x −y) +y

sin(3x −y) = y +xy

y

(sin(3x −y) −x) = y −1+3sin(3x −y)
y

=
y −1+3sin(3x −y)
sin(3x −y) −x
June 8, 2011 LTSV SSM Second Pass
150 C HA P T E R 3 DIFFERENTIATION
2x
2
−x −y =
_
x
4
+y
4
27. Showthat x + yx
−1
= 1andy = x − x
2
deﬁnethesamecurve(except that (0, 0) is not asolution of theﬁrst
equation) andthat implicit differentiationyieldsy

= yx
−1
−x andy

= 1−2x. Explainwhytheseformulasproduce
thesamevaluesfor thederivative.
solution Multiplytheﬁrst equationbyx andthenisolatethey termtoobtain
x
2
+y = x ⇒ y = x −x
2
.
Implicit differentiationappliedtotheﬁrst equationyields
1−yx
−2
+x
−1
y

= 0 or y

= yx
−1
−x.
Fromtheﬁrst equation, weﬁndyx
−1
= 1−x; uponsubstitutingthisexpressionintothepreviousderivative, weﬁnd
y

= 1−x −x = 1−2x,
whichisthederivativeof thesecondequation.
Usethemethodof Example4tocompute
dy
dx
¸
¸
P
at P = (2, 1) onthecurvey
2
x
3
+y
3
x
4
−10x +y = 5.
In Exercises 29 and 30, ﬁnd dy/dx at the given point.
29. (x +2)
2
−6(2y +3)
2
= 3, (1, −1)
solution Bythescalingandshiftingrule,
2(x +2) −24(2y +3)y

= 0.
If x = 1andy = −1, then
2(3) −24(1)y

= 0.
sothat 24y

= 6, or y

=
1
4
.
sin
2
(3y) = x +y,
_
2−π
4
,
π
4
_
In Exercises 31–38, ﬁnd an equation of the tangent line at the given point.
31. xy +x
2
y
2
= 5, (2, 1)
solution Takingthederivativeof bothsidesof xy +x
2
y
2
= 5yields
xy

+y +2xy
2
+2x
2
yy

= 0.
Substitutingx = 2, y = 1, weﬁnd
2y

+1+4+8y

= 0 or y

= −
1
2
.
Hence, theequationof thetangent lineat (2, 1) isy −1= −
1
2
(x −2) or y = −
1
2
x +2.
x
2/3
+y
2/3
= 2, (1, 1)
33. x
2
+siny = xy
2
+1, (1, 0)
solution Takingthederivativeof bothsidesof x
2
+siny = xy
2
+1yields
2x +cosyy

= y
2
+2xyy

.
Substitutingx = 1, y = 0, weﬁnd
2+y

= 0 or y

= −2.
Hence, theequationof thetangent lineisy −0= −2(x −1) or y = −2x +2.
sin(x −y) = x cos
_
y +
π
4
_
,
_
π
4
,
π
4
_
35. 2x
1/2
+4y
−1/2
= xy, (1, 4)
solution Takingthederivativeof bothsidesof 2x
1/2
+4y
−1/2
= xy yields
x
−1/2
−2y
−3/2
y

= xy

+y.
Substitutingx = 1, y = 4, weﬁnd
1−2
_
1
8
_
y

= y

+4 or y

= −
12
5
.
Hence, theequationof thetangent lineisy −4= −
12
5
(x −1) or y = −
12
5
x +
32
5
.
x
x +1
+
y
y +1
= 1, (1, 1)
June 8, 2011 LTSV SSM Second Pass
S E C T I ON 3.8 Implicit Differentiation 151
37. sin(2x −y) =
x
2
y
, (0, π)
solution Usingimplicit differentiation,
2cos(2x −y) −y

cos(2x −y) =
2x
y

x
2
y

y
2
At (0, π), this gives 2cos(−π) − y

cos(−π) = 0, or −2+ y

= 0, so that y

= 2. Thetangent lineat (0, π) is thus
y −π = 2x, or y = 2x +π.
x +

x = y
2
+y
4
, (1, 1)
39. Findthepointsonthegraphof y
2
= x
3
−3x +1(Figure5) wherethetangent lineishorizontal.
(a) First showthat 2yy

= 3x
2
−3, wherey

= dy/dx.
(b) Donot solvefor y

. Rather, set y

= 0andsolvefor x. Thisyieldstwovaluesof x wheretheslopemaybezero.
(c) Showthat thepositivevalueof x doesnot correspondtoapoint onthegraph.
(d) The negative value corresponds to the two points on the graph where the tangent line is horizontal. Find their
coordinates.
2
−2
−2 −1 1 2
x
y
FIGURE 5 Graphof y
2
= x
3
−3x +1.
solution
(a) Applyingimplicit differentiationtoy
2
= x
3
−3x +1, wehave
2y
dy
dx
= 3x
2
−3.
(b) Settingy

= 0wehave0= 3x
2
−3, sox = 1or x = −1.
(c) If wereturntotheequationy
2
= x
3
−3x +1andsubstitutex = 1, weobtaintheequationy
2
= −1, whichhasno
real solutions.
(d) Substitutingx = −1intoy
2
= x
3
−3x +1yields
y
2
= (−1)
3
−3(−1) +1= −1+3+1= 3,
soy =

3or −

3. Thetangent ishorizontal at thepoints(−1,

3) and(−1, −

3).
Show, bydifferentiatingtheequation, that if thetangent lineat apoint (x, y) onthecurvex
2
y −2x +8y = 2is
horizontal, thenxy = 1. Thensubstitutey = x
−1
inx
2
y −2x +8y = 2toshowthat thetangent lineishorizontal
at thepoints
_
2,
1
2
_
and
_
−4, −
1
4
_
.
41. Findall pointsonthegraphof 3x
2
+4y
2
+3xy = 24wherethetangent lineishorizontal (Figure6).
x
y
FIGURE 6 Graphof 3x
2
+4y
2
+3xy = 24.
solution Differentiatingtheequation3x
2
+4y
2
+3xy = 24implicitlyyields
6x +8yy

+3xy

+3y = 0,
so
y

= −
6x +3y
8y +3x
.
Settingy

= 0leadsto6x +3y = 0, or y = −2x. Substitutingy = −2x intotheequation3x
2
+4y
2
+3xy = 24yields
3x
2
+4(−2x)
2
+3x(−2x) = 24,
June 8, 2011 LTSV SSM Second Pass
152 C HA P T E R 3 DIFFERENTIATION
or 13x
2
= 24. Thus, x = ±2

78/13, andthecoordinates of thetwo points onthegraphof 3x
2
+ 4y
2
+ 3xy = 24
wherethetangent lineishorizontal are
_
2

78
13
, −
4

78
13
_
and
_

2

78
13
,
4

78
13
_
.
Showthat nopoint onthegraphof x
2
−3xy +y
2
= 1hasahorizontal tangent line.
43. Figure1showsthegraphof y
4
+xy = x
3
−x +2. Finddy/dx at thetwopointsonthegraphwithx-coordinate0
andﬁndanequationof thetangent lineat (1, 1).
solution Consider theequationy
4
+xy = x
3
−x +2. Then4y
3
y

+xy

+y = 3x
2
−1, and
y

=
3x
2
−y −1
x +4y
3
.
• Substitutingx = 0into y
4
+ xy = x
3
− x + 2gives y
4
= 2, whichhas two real solutions, y = ±2
1/4
. When
y = 2
1/4
, wehave
y

=
−2
1/4
−1
4
_
2
3/4
_ = −

2+
4

2
8
≈ −0.3254.
Wheny = −2
1/4
, wehave
y

=
2
1/4
−1
−4
_
2
3/4
_ = −

2−
4

2
8
≈ −0.02813.
• At thepoint (1, 1), wehavey

=
1
5
. At thispoint thetangent lineisy −1=
1
5
(x −1) or y =
1
5
x +
4
5
.
FoliumofDescartesThecurvex
3
+y
3
= 3xy (Figure7) wasﬁrstdiscussedin1638bytheFrenchphilosopher-
mathematicianRenéDescartes, whocalledit thefolium(meaning“leaf”). Descartes’sscientiﬁccolleagueGillesde
Roberval called it thejasmineﬂower. Both men believed incorrectly that theleaf shapein theﬁrst quadrant was
repeatedineachquadrant, givingtheappearanceof petalsof aﬂower. Findanequationof thetangentlineatthepoint
_
2
3
,
4
3
_
.
45. Findapoint onthefoliumx
3
+y
3
= 3xy other thantheoriginat whichthetangent lineishorizontal.
solution Usingimplicit differentiation, weﬁnd
d
dx
_
x
3
+y
3
_
=
d
dx
(3xy)
3x
2
+3y
2
y

= 3(xy

+y)
Settingy

= 0inthis equationyields 3x
2
= 3y or y = x
2
. If wesubstitutethis expressioninto theoriginal equation
x
3
+y
3
= 3xy, weobtain:
x
3
+x
6
= 3x(x
2
) = 3x
3
or x
3
(x
3
−2) = 0.
Onesolutionof thisequationisx = 0andtheother isx = 2
1/3
. Thus, thetwopointsonthefoliumx
3
+ y
3
= 3xy at
whichthetangent lineishorizontal are(0, 0) and(2
1/3
, 2
2/3
).
Plotx
3
+y
3
= 3xy +b forseveral valuesof b anddescribehowthegraphchangesasb → 0. Thencompute
dy/dx at thepoint (b
1/3
, 0). Howdoesthisvaluechangeasb → ∞? Doyour plotsconﬁrmthisconclusion?
47. Findthex-coordinatesof thepointswherethetangentlineishorizontal onthetrident curve xy = x
3
−5x
2
+2x −1,
sonamedbyIsaacNewtoninhistreatiseoncurvespublishedin1710(Figure8).
Hint: 2x
3
−5x
2
+1= (2x −1)(x
2
−2x −1).
20
−20
−2 8 6
4
2
x
y
FIGURE 8 Trident curve: xy = x
3
−5x
2
+2x −1.
solution Takethederivativeof theequationof atrident curve:
xy = x
3
−5x
2
+2x −1
toobtain
xy

+y = 3x
2
−10x +2.
Settingy

= 0givesy = 3x
2
−10x +2. Substitutingthisintotheequationof thetrident, wehave
xy = x(3x
2
−10x +2) = x
3
−5x
2
+2x −1
June 8, 2011 LTSV SSM Second Pass
S E C T I ON 3.8 Implicit Differentiation 153
or
3x
3
−10x
2
+2x = x
3
−5x
2
+2x −1
Collectingliketermsandsettingtozero, wehave
0= 2x
3
−5x
2
+1= (2x −1)(x
2
−2x −1).
Hence, x =
1
2
, 1±

2.
Findanequationof thetangent lineat eachof thefour pointsonthecurve(x
2
+y
2
−4x)
2
= 2(x
2
+y
2
) where
x = 1. Thiscurve(Figure9) isanexampleof alimaçon of Pascal, namedafter thefather of theFrenchphilosopher
BlaisePascal, whoﬁrst describedit in1650.
49. Findthederivativeat thepointswherex = 1onthefolium(x
2
+y
2
)
2
=
25
4
xy
2
. SeeFigure10.
2
−2
1
x
y
FIGURE 10 Foliumcurve: (x
2
+y
2
)
2
=
25
4
xy
2
solution First, ﬁndthepoints(1, y) onthecurve. Settingx = 1intheequation(x
2
+y
2
)
2
=
25
4
xy
2
yields
(1+y
2
)
2
=
25
4
y
2
y
4
+2y
2
+1=
25
4
y
2
4y
4
+8y
2
+4= 25y
2
4y
4
−17y
2
+4= 0
(4y
2
−1)(y
2
−4) = 0
y
2
=
1
4
or y
2
= 4
Hencey = ±
1
2
or y = ±2. Taking
d
dx
of bothsidesof theoriginal equationyields
2(x
2
+y
2
)(2x +2yy

) =
25
4
y
2
+
25
2
xyy

4(x
2
+y
2
)x +4(x
2
+y
2
)yy

=
25
4
y
2
+
25
2
xyy

(4(x
2
+y
2
) −
25
2
x)yy

=
25
4
y
2
−4(x
2
+y
2
)x
y

=
25
4
y
2
−4(x
2
+y
2
)x
y(4(x
2
+y
2
) −
25
2
x)
• At (1, 2), x
2
+y
2
= 5, and
y

=
25
4
2
2
−4(5)(1)
2(4(5) −
25
2
(1))
=
1
3
.
• At (1, −2), x
2
+y
2
= 5aswell, and
y

=
25
4
(−2)
2
−4(5)(1)
−2(4(5) −
25
2
(1))
= −
1
3
.
• At (1,
1
2
), x
2
+y
2
=
5
4
, and
y

=
25
4
_
1
2
_
2
−4
_
5
4
_
(1)
1
2
_
4
_
5
4
_

25
2
(1)
_ =
11
12
.
June 8, 2011 LTSV SSM Second Pass
154 C HA P T E R 3 DIFFERENTIATION
• At (1, −
1
2
), x
2
+y
2
=
5
4
, and
y

=
25
4
_

1
2
_
2
−4
_
5
4
_
(1)

1
2
_
4
_
5
4
_

25
2
(1)
_ = −
11
12
.
Thefoliumanditstangent linesareplottedbelow:
2
1
−1
−2
2 1.5 1 0.5
x
y
Plot (x
2
+y
2
)
2
= 12(x
2
−y
2
) +2for −4≤ x ≤ 4, 4≤ y ≤ 4usingacomputer algebrasystem. Howmany
horizontal tangent linesdoesthecurveappear tohave? Findthepointswheretheseoccur.
Exercises 51–53: If the derivative dx/dy (instead of dy/dx = 0) exists at a point and dx/dy = 0, then the tangent line
at that point is vertical.
51. Calculatedx/dy fortheequationy
4
+1= y
2
+x
2
andﬁndthepointsonthegraphwherethetangentlineisvertical.
solution Let y
4
+1= y
2
+x
2
. Differentiatingthisequationwithrespect toy yields
4y
3
= 2y +2x
dx
dy
,
so
dx
dy
=
4y
3
−2y
2x
=
y(2y
2
−1)
x
.
Thus,
dx
dy
= 0 when y = 0 and when y = ±

2
2
. Substituting y = 0 into the equation y
4
+ 1 = y
2
+ x
2
gives
1 = x
2
, sox = ±1. Substitutingy = ±

2
2
, givesx
2
= 3/4, sox = ±

3
2
. Thus, therearesix pointsonthegraphof
y
4
+1= y
2
+x
2
wherethetangent lineisvertical:
(1, 0), (−1, 0),
_

3
2
,

2
2
_
,
_

3
2
,

2
2
_
,
_

3
2
, −

2
2
_
,
_

3
2
, −

2
2
_
.
Showthat thetangent linesat x = 1±

2totheconchoid withequation(x −1)
2
(x
2
+y
2
) = 2x
2
arevertical
(Figure11).
53. Useacomputer algebrasystemtoplot y
2
= x
3
−4x for −4≤ x ≤ 4, 4≤ y ≤ 4. Showthat if dx/dy = 0,
theny = 0. Concludethat thetangent lineisvertical at thepointswherethecurveintersectsthex-axis. Doesyour plot
conﬁrmthisconclusion?
solution A plot of thecurvey
2
= x
3
−4x isshownbelow.
1
2
−1
−2
−1 −2 3 2 1
x
y
Differentiatingtheequationy
2
= x
3
−4x withrespect toy yields
2y = 3x
2
dx
dy
−4
dx
dy
,
or
dx
dy
=
2y
3x
2
−4
.
Fromhere, it followsthat
dx
dy
= 0wheny = 0, sothetangent linetothiscurveisvertical at thepointswherethecurve
intersectsthex-axis. Thisconclusionisconﬁrmedbytheplot of thecurveshownabove.
June 8, 2011 LTSV SSM Second Pass
S E C T I ON 3.8 Implicit Differentiation 155
Showthat for all pointsP onthegraphinFigure12, thesegmentsOP andPR haveequal length.
In Exercises 55–58, use implicit differentiation to calculate higher derivatives.
55. Consider theequationy
3

3
2
x
2
= 1.
(a) Showthat y

= x/y
2
anddifferentiateagaintoshowthat
y

=
y
2
−2xyy

y
4
(b) Expressy

intermsof x andy usingpart (a).
solution
(a) Let y
3

3
2
x
2
= 1. Then3y
2
y

−3x = 0, andy

= x/y
2
. Therefore,
y

=
y
2
· 1−x · 2yy

y
4
=
y
2
−2xyy

y
4
.
(b) Substitutingtheexpressionfor y

intotheresult for y

gives
y

=
y
2
−2xy
_
x/y
2
_
y
4
=
y
3
−2x
2
y
5
.
Usethemethodof thepreviousexercisetoshowthat y

= −y
−3
onthecirclex
2
+y
2
= 1.
57. Calculatey

at thepoint (1, 1) onthecurvexy
2
+y −2= 0bythefollowingsteps:
(a) Findy

byimplicit differentiationandcalculatey

at thepoint (1, 1).
(b) Differentiatetheexpressionfor y

foundin(a). Thencomputey

at (1, 1) by substitutingx = 1, y = 1, andthe
valueof y

foundin(a).
solution Let xy
2
+y −2= 0.
(a) Thenx · 2yy

+y
2
· 1+y

= 0, andy

= −
y
2
2xy +1
. At (x, y) = (1, 1), wehavey

= −
1
3
.
(b) Therefore,
y

= −
(2xy +1)
_
2yy

_
−y
2
_
2xy

+2y
_
(2xy +1)
2
= −
(3)
_

2
3
_
−(1)
_

2
3
+2
_
3
2
= −
−6+2−6
27
=
10
27
giventhat (x, y) = (1, 1) andy

= −
1
3
.
Usethemethodof thepreviousexercisetocomputey

at thepoint (1, 1) onthecurvex
3
+y
3
= 3x +y −2.
In Exercises 59–61, x and y are functions of a variable t and use implicit differentiation to relate dy/dt and dx/dt .
59. Differentiatexy = 1withrespect tot andderivetherelation
dy
dt
= −
y
x
dx
dt
.
solution Let xy = 1. Thenx
dy
dt
+y
dx
dt
= 0, and
dy
dt
= −
y
x
dx
dt
.
Differentiatex
3
+3xy
2
= 1withrespect tot andexpressdy/dt intermsof dx/dt , asinExercise59.
61. Calculatedy/dt intermsof dx/dt .
(a) x
3
−y
3
= 1 (b) y
4
+2xy +x
2
= 0
solution
(a) Takingthederivativeof bothsidesof theequationx
3
−y
3
= 1withrespect tot yields
3x
2
dx
dt
−3y
2
dy
dt
= 0 or
dy
dt
=
x
2
y
2
dx
dt
.
(b) Takingthederivativeof bothsidesof theequationy
4
+2xy +x
2
= 0withrespect tot yields
4y
3
dy
dt
+2x
dy
dt
+2y
dx
dt
+2x
dx
dt
= 0,
or
dy
dt
= −
x +y
2y
3
+x
dx
dt
.
ThevolumeV and pressureP of gas in apiston (which vary in timet ) satisfy PV
3/2
= C, whereC is a
constant. Provethat
dP/dt
= −
3 P
June 8, 2011 LTSV SSM Second Pass
156 C HA P T E R 3 DIFFERENTIATION
Further Insights and Challenges
63. Showthat if P liesontheintersectionof thetwocurvesx
2
−y
2
= c andxy = d (c, d constants), thenthetangents
tothecurvesat P areperpendicular.
solution Let C1bethecurvedescribedbyx
2
−y
2
= c, andlet C2bethecurvedescribedbyxy = d. Supposethat
P = (x
0
, y
0
) liesontheintersectionof thetwocurvesx
2
−y
2
= c andxy = d. Sincex
2
−y
2
= c, thechainrulegives
us2x − 2yy

= 0, sothat y

=
2x
2y
=
x
y
. Theslopetothetangent linetoC1is
x
0
y
0
. OnthecurveC2, sincexy = d, the
product ruleyields that xy

+ y = 0, so that y

= −
y
x
. Thereforetheslopeto thetangent lineto C2is −
y
0
x
0
. Thetwo
slopesarenegativereciprocalsof oneanother, hencethetangentstothetwocurvesareperpendicular.
Thelemniscate curve (x
2
+y
2
)
2
= 4(x
2
−y
2
) wasdiscoveredbyJ acobBernoulli in1694, whonotedthat it is
“shapedlikeaﬁgure8, or aknot, or thebowof aribbon.”Findthecoordinatesof thefour pointsatwhichthetangent
lineishorizontal (Figure13).
65. DividethecurveinFigure14
y
5
−y = x
2
y +x +1
intoﬁvebranches, eachof whichisthegraphof afunction. Sketchthebranches.
2
−2
−2 −4 4 2
x
y
FIGURE 14 Graphof y
5
−y = x
2
y +x +1.
solution Thebranchesare:
• Upper branch:
−2 −4 4 2
x
2
−2
y
• Lower part of lower left curve:
x
y
−4 −3 −2 −1
−2
−1
1
• Upper part of lower left curve:
x
y
−4 −3 −2 −1
−1
1
−2
• Upper part of lower right curve:
y
−1
−2
1
1 2 3 4
x
June 8, 2011 LTSV SSM Second Pass
S E C T I ON 3.9 Related Rates 157
• Lower part of lower right curve:
y
−1
−2
1
1 2 3 4
x
3.9 Related Rates
Preliminary Questions
1. Assignvariablesandrestatethefollowingproblemintermsof knownandunknownderivatives(butdonotsolveit):
Howfast isthevolumeof acubeincreasingif itssideincreasesat arateof 0.5cm/s?
solution Let s andV denotethelengthof thesideandthecorrespondingvolumeof acube, respectively. Determine
dV
dt
if
ds
dt
= 0.5cm/s.
2. What istherelationbetweendV/dt anddr/dt if V =
_
4
3
_
πr
3
?
solution Applyingthegeneral power rule, weﬁnd
dV
dt
= 4πr
2dr
dt
. Therefore, theratiois4πr
2
.
In Questions 3 and 4, water pours into a cylindrical glass of radius 4 cm. Let V and h denote the volume and water level
respectively, at time t .
3. Restatethisquestionintermsof dV/dt anddh/dt : Howfast isthewater level risingif water poursinat arateof
2cm
3
/min?
solution Determine
dh
dt
if
dV
dt
= 2cm
3
/min.
4. Restatethisquestionintermsof dV/dt anddh/dt : At what rateiswater pouringinif thewater level risesat arate
of 1cm/min?
solution Determine
dV
dt
if
dh
dt
= 1cm/min.
Exercises
In Exercises 1 and 2, consider a rectangular bathtub whose base is 18 ft
2
.
1. Howfast isthewater level risingif water isﬁllingthetubat arateof 0.7ft
3
/min?
solution Leth betheheightof thewaterinthetubandV bethevolumeof thewater.ThenV = 18h and
dV
dt
= 18
dh
dt
.
Thus
dh
dt
=
1
18
dV
dt
=
1
18
(0.7) ≈ 0.039ft/min.
At what rateiswater pouringintothetubif thewater level risesat arateof 0.8ft/min?
3. Theradiusof acircular oil slickexpandsat arateof 2m/min.
(a) Howfast istheareaof theoil slickincreasingwhentheradiusis25m?
(b) If theradiusis0at timet = 0, howfast istheareaincreasingafter 3min?
solution Let r betheradiusof theoil slickandA itsarea.
(a) ThenA = πr
2
and
dA
dt
= 2πr
dr
dt
. Substitutingr = 25and
dr
dt
= 2, weﬁnd
dA
dt
= 2π (25) (2) = 100π ≈ 314.16m
2
/min.
(b) Since
dr
dt
= 2andr(0) = 0, it followsthat r(t ) = 2t . Thus, r(3) = 6and
dA
dt
= 2π (6) (2) = 24π ≈ 75.40m
2
/min.
June 8, 2011 LTSV SSM Second Pass
158 C HA P T E R 3 DIFFERENTIATION
At what rateisthediagonal of acubeincreasingif itsedgesareincreasingat arateof 2cm/s?
In Exercises 5–8, assume that the radius r of a sphere is expanding at a rate of 30cm/min. The volume of a sphere is
V =
4
3
πr
3
and its surface area is 4πr
2
. Determine the given rate.
5. Volumewithrespect totimewhenr = 15cm.
solution Astheradiusisexpandingat 30centimetersper minute, weknowthat
dr
dt
= 30cm/min. Taking
d
dt
of the
equationV =
4
3
πr
3
yields
dV
dt
=
4
3
π
_
3r
2
dr
dt
_
= 4πr
2
dr
dt
.
Substitutingr = 15and
dr
dt
= 30yields
dV
dt
= 4π(15)
2
(30) = 27,000π cm
3
/min.
Volumewithrespect totimeat t = 2min, assumingthat r = 0at t = 0.
7. Surfaceareawithrespect totimewhenr = 40cm.
solution Takingthederivativeof bothsidesof A = 4πr
2
withrespect tot yields
dA
dt
= 8πr
dr
dt
.
dr
dt
= 30, so
dA
dt
= 8π(40)(30) = 9600π cm
2
/min.
Surfaceareawithrespect totimeat t = 2min, assumingthat r = 10at t = 0.
In Exercises 9–12, refer to a 5-meter ladder sliding down a wall, as in Figures 1 and 2. The variable h is the height of the
ladder’s top at time t , and x is the distance from the wall to the ladder’s bottom.
9. Assumethebottomslidesawayfromthewall at arateof 0.8m/s. Findthevelocityof thetopof theladder at t = 2s
if thebottomis1.5mfromthewall at t = 0s.
solution Let x denotethedistancefromthebaseof theladder tothewall, andh denotetheheight of thetopof the
ladder fromtheﬂoor. Theladder is5mlong, soh
2
+x
2
= 5
2
. At anytimet , x = 1.5+0.8t . Therefore, at timet = 2,
thebaseisx = 1.5+0.8(2) = 3.1mfromthewall. Furthermore, wehave
2h
dh
dt
+2x
dx
dt
= 0 so
dh
dt
= −
x
h
dx
dt
.
Substitutingx = 3.1, h =
_
5
2
−3.1
2
and
dx
dt
= 0.8, weobtain
dh
dt
= −
3.1
_
5
2
−3.1
2
(0.8) ≈ −0.632m/s.
Supposethat thetopisslidingdownthewall at arateof 1.2m/s. Calculatedx/dt whenh = 3m.
11. Supposethat h(0) = 4andthetopslidesdownthewall at arateof 1.2m/s. Calculatex anddx/dt at t = 2s.
solution Lethandx betheheightof theladder’stopandthedistancefromthewall of theladder’sbottom,respectively.
After 2seconds, h = 4+2(−1.2) = 1.6m. Sinceh
2
+x
2
= 5
2
,
x =
_
5
2
−1.6
2
= 4.737m.
Furthermore, wehave2h
dh
dt
+2x
dx
dt
= 0, sothat
dx
dt
= −
h
x
dh
dt
. Substitutingh = 1.6, x = 4.737, and
dh
dt
= −1.2, we
ﬁnd
dx
dt
= −
1.6
4.737
(−1.2) ≈ 0.405m/s.
What is therelationbetweenh andx at themoment whenthetopandbottomof theladder moveat thesame
speed?
13. A conical tank hasheight 3mandradius2mat thetop. Water ﬂowsinat arateof 2m
3
/min. Howfast isthewater
level risingwhenit is2m?
solution Consider theconeof water inthetank at acertaininstant. Let r betheradius of its (inverted) base, h its
height, andV itsvolume. Bysimilar triangles,
r
h
=
2
3
or r =
2
3
h andthusV =
1
3
πr
2
h =
4
27
πh
3
. Therefore,
dV
dt
=
4
9
πh
2
dh
dt
,
and
dh
dt
=
9
4πh
2
dV
dt
.
Substitutingh = 2and
dV
dt
= 2yields
dh
dt
=
9
4π (2)
2
×2=
9

≈ −0.36m/min.
June 8, 2011 LTSV SSM Second Pass
S E C T I ON 3.9 Related Rates 159
Followthesameset-upasExercise13, but assumethat thewater level isrisingat arateof 0.3m/minwhenit is
2m. At what rateiswater ﬂowingin?
15. Theradiusr andheighth of acircular conechangeatarateof 2cm/s. Howfastisthevolumeof theconeincreasing
whenr = 10andh = 20?
solution Let r betheradius, h betheheight, andV bethevolumeof aright circular cone. ThenV =
1
3
πr
2
h, and
dV
dt
=
1
3
π
_
r
2
dh
dt
+2hr
dr
dt
_
.
Whenr = 10, h = 20, and
dr
dt
=
dh
dt
= 2, weﬁnd
dV
dt
=
π
3
_
10
2
· 2+2· 20· 10· 2
_
=
1000π
3
≈ 1047.20cm
3
/s.
Aroadperpendicular toahighwayleadstoafarmhouselocated2kmaway(Figure8).Anautomobiletravelspast
thefarmhouseat aspeedof 80km/h. Howfast isthedistancebetweentheautomobileandthefarmhouseincreasing
whentheautomobileis6kmpast theintersectionof thehighwayandtheroad?
17. A manof height 1.8meters walks away froma5-meter lamppost at aspeedof 1.2m/s (Figure9). Findtherateat
whichhisshadowisincreasinginlength.
x y
5
FIGURE 9
solution Sincethemanismovingat arateof 1.2m/s, hisdistancefromthelight post at anygiventimeisx = 1.2t .
Knowingthemanis1.8meterstall andthat thelengthof hisshadowisdenotedby y, weset upaproportionof similar
trianglesfromthediagram:
y
1.8
=
1.2t +y
5
.
Clearingfractionsandsolvingfor y yields
y = 0.675t.
Thus, dy/dt = 0.675metersper secondistherateat whichthelengthof theshadowisincreasing.
AsClaudiawalksawayfroma264-cmlamppost, thetipof her shadowmovestwiceasfast asshedoes. What is
Claudia’sheight?
19. At agivenmoment, aplanepassesdirectlyabovearadar stationat analtitudeof 6km.
(a) Theplane’s speedis 800km/h. Howfast is thedistancebetweentheplaneandthestationchanginghalf aminute
later?
(b) Howfastisthedistancebetweentheplaneandthestationchangingwhentheplanepassesdirectlyabovethestation?
solution Let x bethedistanceof theplanefromthestationalongthegroundandh thedistancethroughtheair.
(a) BythePythagoreanTheorem, wehave
h
2
= x
2
+6
2
= x
2
+36.
Thus2h
dh
dt
= 2x
dx
dt
, and
dh
dt
=
x
h
dx
dt
. After half aminute, x =
1
2
×
1
60
×800=
20
3
kilometers. Withx =
20
3
,
h =
_
_
20
3
_
2
+36=
1
3

724=
2
3

181≈ 8.969km,
and
dx
dt
= 800,
dh
dt
=
20
3
3
2

181
×800=
8000

181
≈ 594.64km/h.
(b) Whentheplaneisdirectlyabovethestation, x = 0, sothedistancebetweentheplaneandthestationisnotchanging,
for at thisinstant wehave
dh
dt
=
0
6
×800= 0km/h.
June 8, 2011 LTSV SSM Second Pass
160 C HA P T E R 3 DIFFERENTIATION
Inthesettingof Exercise19, letθ betheanglethatthelinethroughtheradar stationandtheplanemakeswiththe
horizontal. Howfast isθ changing12minafter theplanepassesover theradar station?
21. A hotair balloonrisingverticallyistrackedbyanobserver located4kmfromthelift-off point. Atacertainmoment,
theanglebetweentheobserver’slineof sightandthehorizontal is
π
5
, anditischangingatarateof 0.2rad/min. Howfast
istheballoonrisingat thismoment?
solution Lety betheheightof theballoon(inmiles) andθ theanglebetweentheline-of-sightandthehorizontal. Via
trigonometry, wehavetanθ =
y
4
. Therefore,
sec
2
θ ·

dt
=
1
4
dy
dt
,
and
dy
dt
= 4

dt
sec
2
θ.
Using

dt
= 0.2andθ =
π
5
yields
dy
dt
= 4(0.2)
1
cos
2
(π/5)
≈ 1.22km/min.
A laser pointer isplacedonaplatformthatrotatesatarateof 20revolutionsper minute. Thebeamhitsawall 8m
away, producingadot of light that moveshorizontallyalongthewall. Let θ betheanglebetweenthebeamandthe
linethroughthesearchlight perpendicular tothewall (Figure10). Howfast isthisdot movingwhenθ =
π
6
?
23. A rocket travelsverticallyat aspeedof 1200km/h. Therocket istrackedthroughatelescopebyanobserver located
16kmfromthelaunchingpad. Findtherateat whichtheanglebetweenthetelescopeandthegroundisincreasing3min
after lift-off.
solution Let y betheheight of therocket andθ theanglebetweenthetelescopeandtheground. Usingtrigonometry,
wehavetanθ =
y
16
. Therefore,
sec
2
θ ·

dt
=
1
16
dy
dt
,
and

dt
=
cos
2
θ
16
dy
dt
.
Aftertherockethastraveledfor3minutes(or
1
20
hour), itsheightis
1
20
×1200= 60km.Atthisinstant, tanθ = 60/16=
15/4andthus
cosθ =
4
_
15
2
+4
2
=
4

241
.
Finally,

dt
=
16/241
16
(1200) =
1200
241
≈ 4.98rad/hr.
Usingatelescope, youtrack arocket that waslaunched4kmaway, recordingtheangleθ betweenthetelescope
andthegroundat half-secondintervals. Estimatethevelocityof therocket if θ(10) = 0.205andθ(10.5) = 0.225.
25. A policecar travelingsouthtowardSioux Falls at 160km/hpursues atruck travelingeast away fromSioux Falls,
Iowa, at 140km/h(Figure11). At timet = 0, thepolicecar is20kmnorthandthetruck is30kmeast of Sioux Falls.
Calculatetherateat whichthedistancebetweenthevehiclesischanging:
(a) At timet = 0
(b) 5minuteslater
160 km/h
140 km/h
Sioux Falls
x
y
FIGURE 11
solution Let y denotethedistancethepolicecar isnorthof SiouxFallsandx thedistancethetruck iseast of Sioux
Falls. Theny = 20−160t andx = 30+140t . If denotesthedistancebetweenthepolicecar andthetruck, then

2
= x
2
+y
2
= (30+140t )
2
+(20−160t )
2
June 8, 2011 LTSV SSM Second Pass
S E C T I ON 3.9 Related Rates 161
and

d
dt
= 140(30+140t ) −160(20−160t ) = 1000+45,200t.
(a) At t = 0, =
_
30
2
+20
2
= 10

13, so
d
dt
=
1000
10

13
=
100

13
13
≈ 27.735km/h.
(b) At t = 5minutes=
1
12
hour,
=
_
_
30+140·
1
12
_
2
+
_
20−160·
1
12
_
2
≈ 42.197km,
and
d
dt
=
1000+45,200·
1
12
42.197
≈ 112.962km/h.
A car travelsdownahighwayat 25m/s. Anobserver stands150mfromthehighway.
(a) Howfast is thedistancefromtheobserver to thecar increasingwhenthecar passes infront of theobserver?
Explainyour answer without makinganycalculations.
(b) Howfast isthedistanceincreasing20slater?
27. Inthesettingof Example5, at acertainmoment, thetractor’sspeedis3m/sandthebaleisrisingat 2m/s. Howfar
isthetractor fromthebaleat thismoment?
solution FromExample5, wehavetheequation
x
dx
dt
_
x
2
+4.5
2
=
dh
dt
,
wherex denotethedistancefromthetractor tothebaleandh denotestheheight of thebale. Given
dx
dt
= 3 and
dh
dt
= 2,
it followsthat
3x
_
4.5
2
+x
2
= 2,
whichyieldsx =

16.2≈ 4.025m.
Placidopullsaropeattachedtoawagonthroughapulleyat arateof q m/s. WithdimensionsasinFigure12:
(a) Findaformulafor thespeedof thewagonintermsof q andthevariablex intheﬁgure.
(b) Findthespeedof thewagonwhenx = 0.6if q = 0.5m/s.
29. J ulianisjoggingaroundacircular trackof radius50m. Inacoordinatesystemwithoriginat thecenter of thetrack,
J ulian’sx-coordinateischangingat arateof −1.25m/swhenhiscoordinatesare(40, 30). Finddy/dt at thismoment.
solution Wehavex
2
+y
2
= 50
2
, so
2x
dx
dt
+2y
dy
dt
= 0 or
dy
dt
= −
x
y
dx
dt
.
Givenx = 40, y = 30anddx/dt = −1.25, weﬁnd
dy
dt
= −
40
30
(−1.25) =
5
3
m/s.
A particlemovescounterclockwisearoundtheellipsewithequation9x
2
+16y
2
= 25(Figure13).
(a) Inwhichof thefour quadrantsisdx/dt > 0? Explain.
(b) Findarelationbetweendx/dt anddy/dt .
(c) Atwhatrateisthex-coordinatechangingwhentheparticlepassesthepoint(1, 1) if itsy-coordinateisincreasing
at arateof 6m/s?
(d) Finddy/dt whentheparticleisat thetopandbottomof theellipse.
In Exercises 31 and 32, assume that the pressure P (in kilopascals) and volume V (in cubic centimeters) of an expanding
gas are related by PV
b
= C, where b and C are constants (this holds in an adiabatic expansion, without heat gain or
loss).
31. FinddP/dt if b = 1.2, P = 8kPa, V = 100cm
2
, anddV/dt = 20cm
3
/min.
solution Let PV
b
= C. Then
PbV
b−1
dV
dt
+V
b
dP
dt
= 0,
and
dP
dt
= −
Pb
V
dV
dt
.
Substitutingb = 1.2, P = 8, V = 100, and
dV
dt
= 20, weﬁnd
dP
dt
= −
(8) (1.2)
100
(20) = −1.92kPa/min.
June 8, 2011 LTSV SSM Second Pass
162 C HA P T E R 3 DIFFERENTIATION
Findb if P = 25kPa, dP/dt = 12kPa/min, V = 100cm
2
, anddV/dt = 20cm
3
/min.
33. Thebasex of therighttriangleinFigure14increasesatarateof 5cm/s, whiletheheightremainsconstantath = 20.
Howfast istheangleθ changingwhenx = 20?
x
q
20
FIGURE 14
solution Wehavecotθ =
x
20
, fromwhich
−csc
2
θ ·

dt
=
1
20
dx
dt
andthus

dt
= −
sin
2
θ
20
dx
dt
.
Wearegiven
dx
dt
= 5andwhenx = h = 20, θ =
π
4
. Hence,

dt
= −
sin
2
_
π
4
_
20
(5) = −
1
8
rad/s.
Twoparallel paths15mapart runeast-west throughthewoods. Brookejogseast ononepathat 10km/h, while
J amail walkswest ontheother pathat 6km/h. If theypasseachother at timet = 0, howfar apart arethey3slater,
andhowfast isthedistancebetweenthemchangingat that moment?
35. A particletravelsalongacurvey = f (x) asinFigure15. Let L(t ) betheparticle’sdistancefromtheorigin.
(a) Showthat
dL
dt
=
_
x +f (x)f

(x)
_
x
2
+f (x)
2
_
dx
dt
if theparticle’slocationat timet isP = (x, f (x)).
(b) CalculateL

(t ) whenx = 1andx = 2if f (x) =
_
3x
2
−8x +9anddx/dt = 4.
x
y
y = f (x)
O
P
θ
1 2
2
FIGURE 15
solution
(a) If theparticle’slocationat timet isP = (x, f (x)), then
L(t ) =
_
x
2
+f (x)
2
.
Thus,
dL
dt
=
1
2
(x
2
+f (x)
2
)
−1/2
_
2x
dx
dt
+2f (x)f

(x)
dx
dt
_
=
_
x +f (x)f

(x)
_
x
2
+f (x)
2
_
dx
dt
.
(b) Givenf (x) =
_
3x
2
−8x +9, it followsthat
f

(x) =
3x −4
_
3x
2
−8x +9
.
Let’sstart withx = 1. Thenf (1) = 2, f

(1) = −
1
2
and
dL
dt
=
_
1−1
_
1
2
+2
2
_
(4) = 0.
Withx = 2, f (2) =

5, f

(2) = 2/

5and
dL
dt
=
2+2
_
2
2
+

5
2
(4) =
16
3
.
June 8, 2011 LTSV SSM Second Pass
S E C T I ON 3.9 Related Rates 163
Let θ betheangleinFigure15, whereP = (x, f (x)). Inthesettingof thepreviousexercise, showthat

dt
=
_
xf

(x) −f (x)
x
2
+f (x)
2
_
dx
dt
Hint: Differentiatetanθ = f (x)/x andobservethat cosθ = x/
_
x
2
+f (x)
2
.
Exercises 37 and 38 refer to the baseball diamond (a square of side 90ft) in Figure 16.
20 ft/s
15 ft/s
s
90 ft
First base
Second base
Homeplate
FIGURE 16
37. Abaseball player runsfromhomeplatetowardﬁrstbaseat20ft/s. Howfastistheplayer’sdistancefromsecondbase
changingwhentheplayer ishalfwaytoﬁrst base?
solution Let x bethedistanceof theplayer fromhomeplateandh theplayer’sdistancefromsecondbase. Usingthe
Pythagoreantheorem, wehaveh
2
= 90
2
+(90−x)
2
. Therefore,
2h
dh
dt
= 2(90−x)
_

dx
dt
_
,
and
dh
dt
= −
90−x
h
dx
dt
.
Wearegiven
dx
dt
= 20. Whentheplayer ishalfwaytoﬁrst base, x = 45andh =
_
90
2
+45
2
, so
dh
dt
= −
45
_
90
2
+45
2
(20) = −4

5≈ −8.94ft/s.
Player 1runs toﬁrst baseat aspeedof 20ft/s whilePlayer 2runs fromsecondbasetothirdbaseat aspeedof
15ft/s. Lets bethedistancebetweenthetwoplayers. Howfastiss changingwhenPlayer 1is30ftfromhomeplate
andPlayer 2is60ft fromsecondbase?
39. Theconical wateringpail inFigure17hasagridof holes. Water ﬂowsout throughtheholesat arateof kA m
3
/min,
where k is a constant and A is the surface area of the part of the cone in contact with the water. This surface area
is A = πr
_
h
2
+r
2
and the volume is V =
1
3
πr
2
h. Calculate the rate dh/dt at which the water level changes at
h = 0.3m, assumingthat k = 0.25m.
0.45 m
0.15 m
h
r
FIGURE 17
solution Bysimilar triangles, wehave
r
h
=
0.15
0.45
=
1
3
so r =
1
3
h.
Substitutingthisexpressionfor r intotheformulafor V yields
V =
1
3
π
_
1
3
h
_
2
h =
1
27
πh
3
.
Fromhereandtheproblemstatement, it followsthat
dV
dt
=
1
9
πh
2
dh
dt
= −kA = −0.25πr
_
h
2
+r
2
.
June 8, 2011 LTSV SSM Second Pass
164 C HA P T E R 3 DIFFERENTIATION
Solvingfor dh/dt gives
dh
dt
= −
9
4
r
h
2
_
h
2
+r
2
.
Whenh = 0.3, r = 0.1and
dh
dt
= −
9
4
0.1
0.3
2
_
0.3
2
+0.1
2
= −0.79m/min.
Further Insights and Challenges
A bowl contains water that evaporates at arateproportional to thesurfaceareaof water exposed to theair
(Figure18). Let A(h) bethecross-sectional areaof thebowl at height h.
(a) ExplainwhyV(h +h) −V(h) ≈ A(h)h if h issmall.
(b) Use(a) toarguethat
dV
dh
= A(h).
(c) Showthat thewater level h decreasesat aconstant rate.
41. A roller coaster hastheshapeof thegraphinFigure19. Showthatwhentheroller coaster passesthepoint (x, f (x)),
thevertical velocityof theroller coaster isequal tof

(x) timesitshorizontal velocity.
(x, f (x))
FIGURE 19 Graphof f (x) asaroller coaster track.
solution Let theequationy = f (x) describetheshapeof theroller coaster track. Taking
d
dt
of bothsides of this
equationyields
dy
dt
= f

(x)
dx
dt
. Inother words, thevertical velocityof acar movingalongthetrack,
dy
dt
, isequal tof

(x)
timesthehorizontal velocity,
dx
dt
.
Twotrainsleaveastationat t = 0andtravel withconstant velocityv alongstraight tracksthat makeanangleθ.
(a) Showthat thetrainsareseparatingfromeachother at aratev

2−2cosθ.
(b) What doesthisformulagivefor θ = π?
43. As thewheel of radius r cminFigure20rotates, therodof lengthL attachedat point P drives apistonback and
forthinastraight line. Let x bethedistancefromtheorigintopoint Qat theendof therod, asshownintheﬁgure.
(a) UsethePythagoreanTheoremtoshowthat
L
2
= (x −r cosθ)
2
+r
2
sin
2
θ 6
(b) DifferentiateEq. (6) withrespect tot toprovethat
2(x −r cosθ)
_
dx
dt
+r sinθ

dt
_
+2r
2
sinθ cosθ

dt
= 0
(c) Calculatethespeedof thepistonwhenθ =
π
2
, assumingthat r = 10cm, L = 30cm, andthewheel rotates at 4
revolutionsper minute.
Piston moves
back and forth
x
L
q
P
Q
r
FIGURE 20
solution Fromthediagram, thecoordinatesof P are(r cosθ, r sinθ) andthoseof Qare(x, 0).
(a) Thedistanceformulagives
L =
_
(x −r cosθ)
2
+(−r sinθ)
2
.
Thus,
L
2
= (x −r cosθ)
2
+r
2
sin
2
θ.
Notethat L (thelengthof theﬁxedrod) andr (theradiusof thewheel) areconstants.
June 8, 2011 LTSV SSM Second Pass
Chapter Review Exercises 165
(b) From(a) wehave
0= 2(x −r cosθ)
_
dx
dt
+r sinθ

dt
_
+2r
2
sinθ cosθ

dt
.
(c) Solvingfor dx/dt in(b) gives
dx
dt
=
r
2
sinθ cosθ

dt
r cosθ −x
−r sinθ

dt
=
rx sinθ

dt
r cosθ −x
.
Withθ =
π
2
, r = 10, L = 30, and

dt
= 8π,
dx
dt
=
(10) (x)
_
sin
π
2
_
(8π)
(10) (0) −x
= −80π ≈ −251.33cm/min
A spectator seated300mawayfromthecenter of acircular trackof radius100mwatchesanathleterunlapsat a
speedof 5m/s. Howfast isthedistancebetweenthespectator andathletechangingwhentherunner isapproaching
thespectator andthedistancebetweenthemis 250m? Hint: Thediagramfor this problemis similar toFigure20,
withr = 100andx = 300.
CHAPTER REVIEW EXERCISES
In Exercises 1–4, refer to the function f (x) whose graph is shown in Figure 1.
y
2.0 1.5 1.0 0.5
x
7
6
5
4
3
2
1
FIGURE 1
1. Computetheaveragerateof changeof f (x) over [0, 2]. What isthegraphical interpretationof thisaveragerate?
solution Theaveragerateof changeof f (x) over [0, 2] is
f (2) −f (0)
2−0
=
7−1
2−0
= 3.
Graphically, thisaveragerateof changerepresentstheslopeof thesecant linethroughthepoints(2, 7) and(0, 1) onthe
graphof f (x).
For which valueof h is
f (0.7+h) −f (0.7)
h
equal to theslopeof thesecant linebetween thepoints where
x = 0.7andx = 1.1?
3. Estimate
f (0.7+h) −f (0.7)
h
for h = 0.3. Isthisnumber larger or smaller thanf

(0.7)?
solution For h = 0.3,
f (0.7+h) −f (0.7)
h
=
f (1) −f (0.7)
0.3

2.8−2
0.3
=
8
3
.
Becausethecurveisconcaveup, theslopeof thesecant lineislarger thantheslopeof thetangent line, sothevalueof
thedifferencequotient shouldbelarger thanthevalueof thederivative.
Estimatef

(0.7) andf

(1.1).
In Exercises 5–8, compute f

(a) using the limit deﬁnition and ﬁnd an equation of the tangent line to the graph of f (x)
at x = a.
5. f (x) = x
2
−x, a = 1
solution Let f (x) = x
2
−x anda = 1. Then
f

(a) = lim
h→0
f (a +h) −f (a)
h
= lim
h→0
(1+h)
2
−(1+h) −(1
2
−1)
h
= lim
h→0
1+2h +h
2
−1−h
h
= lim
h→0
(1+h) = 1
andtheequationof thetangent linetothegraphof f (x) at x = a is
y = f

(a)(x −a) +f (a) = 1(x −1) +0= x −1.
June 8, 2011 LTSV SSM Second Pass
166 C HA P T E R 3 DIFFERENTIATION
f (x) = 5−3x, a = 2
7. f (x) = x
−1
, a = 4
solution Let f (x) = x
−1
anda = 4. Then
f

(a) = lim
h→0
f (a +h) −f (a)
h
= lim
h→0
1
4+h

1
4
h
= lim
h→0
4−(4+h)
4h(4+h)
= lim
h→0
−1
4(4+h)
= −
1
4(4+0)
= −
1
16
andtheequationof thetangent linetothegraphof f (x) at x = a is
y = f

(a)(x −a) +f (a) = −
1
16
(x −4) +
1
4
= −
1
16
x +
1
2
.
f (x) = x
3
, a = −2
In Exercises 9–12, compute dy/dx using the limit deﬁnition.
9. y = 4−x
2
solution Let y = 4−x
2
. Then
dy
dx
= lim
h→0
4−(x +h)
2
−(4−x
2
)
h
= lim
h→0
4−x
2
−2xh −h
2
−4+x
2
h
= lim
h→0
(−2x −h) = −2x −0= −2x.
y =

2x +1
11. y =
1
2−x
solution Let y =
1
2−x
. Then
dy
dx
= lim
h→0
1
2−(x+h)

1
2−x
h
= lim
h→0
(2−x) −(2−x −h)
h(2−x −h)(2−x)
= lim
h→0
1
(2−x −h)(2−x)
=
1
(2−x)
2
.
y =
1
(x −1)
2
In Exercises 13–16, express the limit as a derivative.
13. lim
h→0

1+h −1
h
solution Let f (x) =

x. Then
lim
h→0

1+h −1
h
= lim
h→0
f (1+h) −f (1)
h
= f

(1).
lim
x→−1
x
3
+1
x +1
15. lim
t →π
sint cost
t −π
solution Let f (t ) = sint cost andnotethat f (π) = sinπ cosπ = 0. Then
lim
t →π
sint cost
t −π
= lim
t →π
f (t ) −f (π)
t −π
= f

(π).
lim
θ→π
cosθ −sinθ +1
θ −π
17. Findf (4) andf

(4) if thetangent linetothegraphof f (x) at x = 4hasequationy = 3x −14.
solution Theequationof thetangent lineto thegraphof f (x) at x = 4is y = f

(4)(x − 4) + f (4) = f

(4)x +
(f (4) −4f

(4)). Matchingthistoy = 3x −14, weseethat f

(4) = 3andf (4) −4(3) = −14, sof (4) = −2.
Each graph in Figure2 shows thegraph of afunction f (x) and its derivativef

(x). Determinewhich is the
functionandwhichisthederivative.
19. Is(A), (B), or (C) thegraphof thederivativeof thefunctionf (x) showninFigure3?
(A) (B)
y
(C)
y
x
−2 2 −1 1
x
−2 2 −1 1
y
y = f (x)
x
−2 2 −1 1
y
x
−2 2 −1 1
FIGURE 3
June 8, 2011 LTSV SSM Second Pass
Chapter Review Exercises 167
solution Thegraphof f (x) hasfour horizontal tangentlineson[−2, 2], sothegraphof itsderivativemusthavefour
x-interceptson[−2, 2]. Thiseliminates(B). Moreover, f (x) isincreasingat bothendsof theinterval, soitsderivative
must bepositiveat bothends. Thiseliminates(A) andidentiﬁes(C) asthegraphof f

(x).
LetN(t ) bethepercentageof astatepopulationinfectedwithaﬂuvirusonweekt of anepidemic.Whatpercentage
islikelytobeinfectedinweek4if N(3) = 8andN

(3) = 1.2?
21. A girl’sheight h(t ) (incentimeters) ismeasuredat timet (inyears) for 0≤ t ≤ 14:
52, 75.1, 87.5, 96.7, 104.5, 111.8, 118.7, 125.2,
131.5, 137.5, 143.3, 149.2, 155.3, 160.8, 164.7
(a) What istheaveragegrowthrateover the14-year period?
(b) Istheaveragegrowthratelarger over theﬁrst half or thesecondhalf of thisperiod?
(c) Estimateh

(t ) (incentimetersper year) for t = 3, 8.
solution
(a) Theaveragegrowthrateover the14-year periodis
164.7−52
14
= 8.05cm/year.
(b) Over theﬁrst half of the14-year period, theaveragegrowthrateis
125.2−52
7
≈ 10.46cm/year,
whichislarger thantheaveragegrowthrateover thesecondhalf of the14-year period:
164.7−125.2
7
≈ 5.64cm/year.
(c) For t = 3,
h

(3) ≈
h(4) −h(3)
4−3
=
104.5−96.7
1
= 7.8cm/year;
for t = 8,
h

(8) ≈
h(9) −h(8)
9−8
=
137.5−131.5
1
= 6.0cm/year.
A planet’s periodP (number of days to completeonerevolutionaroundthesun) is approximately 0.199A
3/2
,
whereA istheaveragedistance(inmillionsof kilometers) fromtheplanet tothesun.
(a) CalculateP anddP/dA for EarthusingthevalueA = 150.
(b) EstimatetheincreaseinP if A isincreasedto152.
In Exercises 23 and 24, use the following table of values for the number A(t ) of automobiles (in millions) manufactured
in the United States in year t .
t 1970 1971 1972 1973 1974 1975 1976
A(t ) 6.55 8.58 8.83 9.67 7.32 6.72 8.50
23. What istheinterpretationof A

(t )? EstimateA

(1971). DoesA

(1974) appear tobepositiveor negative?
solution Because A(t ) measures the number of automobiles manufactured in the United States in year t , A

(t )
measurestherateof changeinautomobileproductionintheUnitedStates. For t = 1971,
A

(1971) ≈
A(1972) −A(1971)
1972−1971
=
8.83−8.58
1
= 0.25millionautomobiles/year.
BecauseA(t ) decreasesfrom1973to1974andfrom1974to1975, it appearsthat A

(1974) wouldbenegative.
Giventhedata, whichof (A)–(C) inFigure4couldbethegraphof thederivativeA

(t )? Explain.
In Exercises 25–50, compute the derivative.
25. y = 3x
5
−7x
2
+4
solution Let y = 3x
5
−7x
2
+4. Then
dy
dx
= 15x
4
−14x.
y = 4x
−3/2
27. y = t
−7.3
solution Let y = t
−7.3
. Then
dy
dt
= −7.3t
−8.3
.
June 8, 2011 LTSV SSM Second Pass
168 C HA P T E R 3 DIFFERENTIATION
y = 4x
2
−x
−2 29. y =
x +1
x
2
+1
solution Let y =
x +1
x
2
+1
. Then
dy
dx
=
(x
2
+1)(1) −(x +1)(2x)
(x
2
+1)
2
=
1−2x −x
2
(x
2
+1)
2
.
y =
3t −2
4t −9
31. y = (x
4
−9x)
6
solution Let y = (x
4
−9x)
6
. Then
dy
dx
= 6(x
4
−9x)
5
d
dx
(x
4
−9x) = 6(4x
3
−9)(x
4
−9x)
5
.
y = (3t
2
+20t
−3
)
6
33. y = (2+9x
2
)
3/2
solution Let y = (2+9x
2
)
3/2
. Then
dy
dx
=
3
2
(2+9x
2
)
1/2
d
dx
(2+9x
2
) = 27x(2+9x
2
)
1/2
.
y = (x +1)
3
(x +4)
4
35. y =
z

1−z
solution Let y =
z

1−z
. Then
dy
dz
=

1−z −(−
z
2
)
1

1−z
1−z
=
1−z +
z
2
(1−z)
3/2
=
2−z
2(1−z)
3/2
.
y =
_
1+
1
x
_
3
37. y =
x
4
+

x
x
2
solution Let
y =
x
4
+

x
x
2
= x
2
+x
−3/2
.
Then
dy
dx
= 2x −
3
2
x
−5/2
.
y =
1
(1−x)

2−x
39. y =
_
x +
_
x +

x
solution Let y =
_
x +
_
x +

x. Then
dy
dx
=
1
2
_
x +
_
x +

x
_
−1/2
d
dx
_
x +
_
x +

x
_
=
1
2
_
x +
_
x +

x
_
−1/2
_
1+
1
2
_
x +

x
_
−1/2
d
dx
_
x +

x
_
_
=
1
2
_
x +
_
x +

x
_
−1/2
_
1+
1
2
_
x +

x
_
−1/2
_
1+
1
2
x
−1/2
__
.
h(z) =
_
z +(z +1)
1/2
_
−3/2
41. y = tan(t
−3
)
solution Let y = tan(t
−3
). Then
dy
dt
= sec
2
(t
−3
)
d
dt
t
−3
= −3t
−4
sec
2
(t
−3
).
y = 4cos(2−3x)
43. y = sin(2x) cos
2
x
solution Let y = sin(2x) cos
2
x = 2sinx cos
3
x. Then
dy
dx
= −6sin
2
x cos
2
x +2cos
4
x.
June 8, 2011 LTSV SSM Second Pass
Chapter Review Exercises 169
y = sin
_
4
θ
_
45. y =
t
1+sect
solution Let y =
t
1+sect
. Then
dy
dt
=
1+sect −t sect tant
(1+sect )
2
.
y = z csc(9z +1) 47. y =
8
1+cotθ
solution Let y =
8
1+cotθ
= 8(1+cotθ)
−1
. Then
dy

= −8(1+cotθ)
−2
d

(1+cotθ) =
8csc
2
θ
(1+cotθ)
2
.
y = tan(cosx)
49. y = tan(

1+cscθ)
solution
dy
dx
= sec
2
(

1+cscθ)
d
dx

1+cscθ
= sec
2
(

1+cscθ) ·
1
2
(1+cscθ)
−1/2
d
dx
(1+cscθ)
= −
sec
2
(

1+cscθ) cscθ cotθ
2(

1+cscθ)
.
y = cos(cos(cos(θ)))
In Exercises 51–56, use the following table of values to calculate the derivative of the given function at x = 2.
x f (x) g(x) f

(x) g

(x)
2 5 4 −3 9
4 3 2 −2 3
51. S(x) = 3f (x) −2g(x)
solution Let S(x) = 3f (x) −2g(x). ThenS

(x) = 3f

(x) −2g

(x) and
S

(2) = 3f

(2) −2g

(2) = 3(−3) −2(9) = −27.
H(x) = f (x)g(x) 53. R(x) =
f (x)
g(x)
solution Let R(x) = f (x)/g(x). Then
R

(x) =
g(x)f

(x) −f (x)g

(x)
g(x)
2
and
R

(2) =
g(2)f

(2) −f (2)g

(2)
g(2)
2
=
4(−3) −5(9)
4
2
= −
57
16
.
G(x) = f (g(x))
55. F(x) = f (g(2x))
solution Let F(x) = f (g(2x)). ThenF

(x) = 2f

(g(2x))g

(2x) and
F

(2) = 2f

(g(4))g

(4) = 2f

(2)g

(4) = 2(−3)(3) = −18.
K(x) = f (x
2
)
57. Findthepointsonthegraphof x
3
−y
3
= 3xy −3wherethetangent lineishorizontal.
solution Useimplicit differentiation:
3x
2
−3y
2
y

= 3y +3xy

3x
2
−3y = y

(3x +3y
2
)
y

=
x
2
−y
x +y
2
June 8, 2011 LTSV SSM Second Pass
170 C HA P T E R 3 DIFFERENTIATION
Thetangent lineis horizontal at thosepoints onthegraphwherey = x
2
. Substitutingx
2
for y intheequationof the
graphgives
x
3
−x
6
= 3x
3
−3, or x
6
+2x
3
−3= 0
Treatingthis as aquadratic equation in x
3
gives x
3
= 1and x
3
= −3. Thus thex-values of thepoints on thecurve
wherethetangent ishorizontal arex = 1andx =
3

−3. Sinceweknowthat at thesepointsy = x
2
, thecorresponding
y-coordinates are1and(−3)
2/3
= 3
2/3
. Thus thecoordinates of thepoints at whichthetangent linetothis graphare
horizontal are(1, 1) and(−3
1/3
, 3
2/3
).
Findthepointsonthegraphof x
2/3
+y
2/3
= 1wherethetangent linehasslope1.
59. Finda suchthat thetangent linesy = x
3
−2x
2
+x +1at x = a andx = a +1areparallel.
solution Let f (x) = x
3
− 2x
2
+ x + 1. Thenf

(x) = 3x
2
− 4x + 1andtheslopeof thetangent lineat x = a is
f

(a) = 3a
2
−4a +1, whiletheslopeof thetangent lineat x = a +1is
f

(a +1) = 3(a +1)
2
−4(a +1) +1= 3(a
2
+2a +1) −4a −4+1= 3a
2
+2a.
Inorder for thetangent linesat x = a andx = a +1tohavethesameslope, wemust havef

(a) = f

(a +1), or
3a
2
−4a +1= 3a
2
+2a.
Theonlysolutiontothisequationisa =
1
6
. Theequationof thetangent lineat x =
1
6
is
y = f

_
1
6
__
x −
1
6
_
+f
_
1
6
_
=
5
12
_
x −
1
6
_
+
241
216
=
5
12
x +
113
108
,
andtheequationof thetangent lineat x =
7
6
is
y = f

_
7
6
__
x −
7
6
_
+f
_
7
6
_
=
5
12
_
x −
7
6
_
+
223
216
=
5
12
x +
59
108
.
Thegraphsof f (x) andthetwotangent linesappear below.
y
x
−1
−1
−3
−2
2 1.5 1 0.5
3
2
In Exercises 60–63, let f (x) = x
3
−3x
2
+x +4.
Findthepointsonthegraphof f (x) wherethetangent linehasslope10.
61. For whichvaluesof x arethetangent linestoy = f (x) horizontal?
solution Thetangent linesarehorizontal whenthederivativevanishes. Fromthepreviousproblem, f

(x) = 3x
2

6x +1, whichiszerowhenx = 1±

6
3
.
Findall valuesof b suchthat y = 25x +b istangent tothegraphof f (x).
63. Findall valuesof k suchthat f (x) hasonlyonetangent lineof slopek.
solution Thetangent linehasslopek whenf

(x) = 3x
2
− 6x + 1 = k, sothat 3x
2
− 6x + (1− k) = 0. Thereis
exactlyonepointwherethetangentlinehasslopek if thisequationhasonlyoneroot, whichoccurswhenitsdiscriminant
iszero. Thediscriminant of thisquadraticis36− 4· 3· (1− k) = 24+ 12k, sothat k = −2. Thusk = −2istheonly
suchvalueof k.
Usethetableto computetheaveragerateof changeof CandidateA’s percentageof votes over theintervals
fromday20today15, day15today10, andday10today5. If thistrendcontinuesover thelast 5daysbeforethe
election, will CandidateA win?
DaysBeforeElection 20 15 10 5
CandidateA 44.8% 46.8% 48.3% 49.3%
CandidateB 55.2% 53.2% 51.7% 50.7%
In Exercises 65–70, calculate y

.
65. y = 12x
3
−5x
2
+3x
solution Let y = 12x
3
−5x
2
+3x. Then
y

= 36x
2
−10x +3 and y

= 72x −10.
y = x
−2/5
67. y =

2x +3
solution Let y =

2x +3= (2x +3)
1/2
. Then
y

=
1
2
(2x +3)
−1/2
d
dx
(2x +3) = (2x +3)
−1/2
and y

= −
1
2
(2x +3)
−3/2
d
dx
(2x +3) = −(2x +3)
−3/2
.
June 8, 2011 LTSV SSM Second Pass
Chapter Review Exercises 171
y =
4x
x +1
69. y = tan(x
2
)
solution Let y = tan(x
2
). Then
y

= 2x sec
2
(x
2
) and
y

= 2x
_
2sec(x
2
)
d
dx
sec(x
2
)
_
+2sec
2
(x
2
) = 8x
2
sec
2
(x
2
) tan(x
2
) +2sec
2
(x
2
).
y = sin
2
(4x +9) In Exercises 71–76, compute
dy
dx
.
71. x
3
−y
3
= 4
solution Consider theequationx
3
−y
3
= 4. Differentiatingwithrespect tox yields
3x
2
−3y
2
dy
dx
= 0.
Therefore,
dy
dx
=
x
2
y
2
.
4x
2
−9y
2
= 36
73. y = xy
2
+2x
2
solution Consider theequationy = xy
2
+2x
2
. Differentiatingwithrespect tox yields
dy
dx
= 2xy
dy
dx
+y
2
+4x.
Therefore,
dy
dx
=
y
2
+4x
1−2xy
.
y
x
= x +y
75. y = sin(x +y)
solution Consider theequationy = sin(x +y). Differentiatingwithrespect tox yields
dy
dx
= cos(x +y)
_
1+
dy
dx
_
.
Therefore,
dy
dx
=
cos(x +y)
1−cos(x +y)
.
tan(x +y) = xy
77. InFigure5, label thegraphsf , f

, andf

.
y
x
y
x
FIGURE 5
solution First consider theplot on theleft. Observethat thegreen curveis nonnegativewhereas thered curveis
increasing, suggestingthat thegreencurveis thederivativeof theredcurve. Moreover, thegreencurveis linear with
negativeslopefor x < 0andlinear withpositiveslopefor x > 0whilethebluecurveisanegativeconstant for x < 0
andapositiveconstant for x > 0, suggestingthebluecurveisthederivativeof thegreencurve. Thus, thered, greenand
bluecurves, respectively, arethegraphsof f , f

andf

.
Nowconsider theplotontheright. Becausetheredcurveisdecreasingwhenthebluecurveisnegativeandincreasing
whenthebluecurveispositiveandthegreencurveisdecreasingwhentheredcurveisnegativeandincreasingwhenthe
redcurveispositive, it followsthat thegreen, redandbluecurves, respectively, arethegraphsof f , f

andf

.
Let f (x) = x
2
sin(x
−1
) for x = 0andf (0) = 0. Showthat f

(x) exists for all x (includingx = 0) but that
f

(x) isnot continuousat x = 0(Figure6).
June 8, 2011 LTSV SSM Second Pass
172 C HA P T E R 3 DIFFERENTIATION
Exercises 79–81: Let q be the number of units of a product (cell phones, barrels of oil, etc.) that can be sold at the price p.
The price elasticity of demand E is deﬁned as the percentage rate of change of q with respect to p. In terms of derivatives,
E =
p
q
dq
dp
= lim
p→0
(100q)/q
(100p)/p
79. Showthat thetotal revenueR = pq satisﬁes
dR
dp
= q(1+E).
solution Let R = pq. Then
dR
dp
= p
dq
dp
+q = q
p
q
dq
dp
+q = q(E +1).
Acommercial bakerycansell q chocolatecakesperweekatprice\$p, whereq = 50p(10−p) for5< p < 10.
(a) Showthat E(p) =
2p −10
p −10
.
(b) Show, bycomputingE(8), that if p = \$8, thena1%increaseinpricereducesdemandbyapproximately3%.
81. Themonthly demand(inthousands) for ﬂights betweenChicago andSt. Louis at thepricep is q = 40− 0.2p.
Calculatethepriceelasticity of demandwhenp = \$150andestimatethepercentageincreaseinnumber of additional
passengersif theticket priceisloweredby1%.
solution Let q = 40−0.2p. Thenq

(p) = −0.2and
E(p) =
_
p
q
_
dq
dp
=
0.2p
0.2p −40
.
For p = 150,
E(150) =
0.2(150)
0.2(150) −40
= −3,
soa1%decreaseinpriceincreasesdemandby 3%. Thedemandwhenp = 150isq = 40− 0.2(150) = 10, or 10,000
passengers. Therefore, a1%increaseindemandtranslatesto300additional passengers.
Howfast doesthewater level riseinthetank inFigure7whenthewater level ish = 4mandwater poursinat
20m
3
/min?
83. Theminutehandof aclockis8cmlong, andthehour handis5cmlong. Howfast isthedistancebetweenthetips
of thehandschangingat 3o’clock?
solution Let S bethedistancebetweenthetipsof thetwohands. Bythelawof cosines
S
2
= 8
2
+5
2
−2· 8· 5cos(θ),
whereθ istheanglebetweenthehands. Thus
2S
dS
dt
= 80sin(θ)

dt
.
At threeo’clockθ = π/2, S =

89, and

dt
=
_
π
360

π
30
_
rad/min= −
11π
360
rad/min,
so
dS
dt
=
1
2

89
(80)(1)
−11π
360
≈ −0.407cm/min.
ChloeandBaoareinmotorboatsat thecenter of alake. At timet = 0, Chloebeginstravelingsouthat aspeedof
50km/h. Oneminutelater, Baotakesoff, headingeast at aspeedof 40km/h. At what rateisthedistancebetween
themincreasingat t = 12min?
85. A beadslides downthecurvexy = 10. Findthebead’s horizontal velocity at timet = 2s if its height at timet
secondsisy = 400−16t
2
cm.
solution Let xy = 10. Thenx = 10/y and
dx
dt
= −
10
y
2
dy
dt
.
If y = 400−16t
2
, then
dy
dt
= −32t and
dx
dt
= −
10
(400−16t
2
)
2
(−32t ) =
320t
(400−16t
2
)
2
.
Thus, at t = 2,
dx
dt
=
640
(336)
2
≈ 0.00567cm/s.
June 8, 2011 LTSV SSM Second Pass
Chapter Review Exercises 173
InFigure8, x isincreasingat2cm/s, y isincreasingat3cm/s, andθ isdecreasingsuchthattheareaof thetriangle
hastheconstant value4cm
2
.
(a) Howfast isθ decreasingwhenx = 4, y = 4?
(b) Howfast isthedistancebetweenP andQchangingwhenx = 4, y = 4?
87. Alightmovingat0.8m/sapproachesamanstanding4mfromawall (Figure9). Thelightis1mabovetheground.
Howfast isthetipP of theman’sshadowmovingwhenthelight is7mfromthewall?
1.8 m
1 m
4 m 0.8 m/s
P
FIGURE 9
solution Let x denotethedistancebetweenthemanandthelight. Usingsimilar triangles, weﬁnd
0.8
x
=
P −1
4+x
or P =
3.2
x
+1.8.
Therefore,
dP
dt
= −
3.2
x
2
dx
dt
.
Whenthelight is7feet fromthewall, x = 3. With
dx
dt
= −0.8, wehave
dP
dt
= −
3.2
3
2
(−0.8) = 0.284m/s.
June 9, 2011 LTSV SSM Second Pass
4 APPLICATIONS OF
THE DERIVATIVE
4.1 Linear Approximation and Applications
Preliminary Questions
1. Trueor False? TheLinear Approximationsays that thevertical changeinthegraphis approximately equal to the
vertical changeinthetangent line.
solution Thisstatementistrue.Thelinearapproximationdoessaythatthevertical changeinthegraphisapproximately
equal tothevertical changeinthetangent line.
2. Estimateg(1.2) −g(1) if g

(1) = 4.
solution UsingtheLinearApproximation,
g(1.2) −g(1) ≈ g

(1)(1.2−1) = 4(0.2) = 0.8.
3. Estimatef (2.1) if f (2) = 1andf

(2) = 3.
solution UsingtheLinearization,
f (2.1) ≈ f (2) +f

(2)(2.1−2) = 1+3(0.1) = 1.3
4. Completethesentence: TheLinearApproximationshowsthatuptoasmall error, thechangeinoutputf isdirectly
proportional to….
solution TheLinearApproximationtellsusthat uptoasmall error, thechangeinoutput f isdirectlyproportional
tothechangeininput x whenx issmall.
Exercises
In Exercises 1–6, use Eq. (1) to estimate f = f (3.02) −f (3).
1. f (x) = x
2
solution Let f (x) = x
2
. Thenf

(x) = 2x andf ≈ f

(3)x = 6(0.02) = 0.12.
f (x) = x
4
3. f (x) = x
−1
solution Let f (x) = x
−1
. Thenf

(x) = −x
−2
andf ≈ f

(3)x = −
1
9
(0.02) = −0.00222.
f (x) =
1
x +1
5. f (x) =

x +6
solution Let f (x) =

x +6. Thenf

(x) =
1
2
(x +6)
−1/2
and
f ≈ f

(3)x =
1
2
9
−1/2
(0.02) = 0.003333.
f (x) = tan
πx
3
7. Thecuberoot of 27is3. Howmuchlarger isthecuberoot of 27.2? EstimateusingtheLinearApproximation.
solution Let f (x) = x
1/3
, a = 27, andx = 0.2. Thenf

(x) =
1
3
x
−2/3
andf

(a) = f

(27) =
1
27
. TheLinear
Approximationis
f ≈ f

(a)x =
1
27
(0.2) = 0.0074074
Estimatesin
_
π
2
+0.1
_
−sin
π
2
usingdifferentials.
In Exercises 9–12, use Eq. (1) to estimate f . Use a calculator to compute both the error and the percentage error.
9. f (x) =

1+x, a = 3, x = 0.2
solution Let f (x) = (1+ x)
1/2
, a = 3, andx = 0.2. Thenf

(x) =
1
2
(1+ x)
−1/2
, f

(a) = f

(3) =
1
4
and
f ≈ f

(a)x =
1
4
(0.2) = 0.05. Theactual changeis
f = f (a +x) −f (a) = f (3.2) −f (3) =

4.2−2≈ 0.049390.
Theerror intheLinearApproximationistherefore|0.049390−0.05| = 0.000610; inpercentageterms, theerror is
0.000610
0.049390
×100%≈ 1.24%.
174
June 9, 2011 LTSV SSM Second Pass
S E C T I ON 4.1 Linear Approximation and Applications 175
f (x) = 2x
2
−x, a = 5, x = −0.4
11. f (x) =
1
1+x
2
, a = 3, x = 0.5
solution Let f (x) =
1
1+x
2
, a = 3, and x = 0.5. Then f

(x) = −
2x
(1+x
2
)
2
, f

(a) = f

(3) = −0.06 and
f ≈ f

(a)x = −0.06(0.5) = −0.03. Theactual changeis
f = f (a +x) −f (a) = f (3.5) −f (3) ≈ −0.0245283.
Theerror intheLinear Approximationis therefore| − 0.0245283− (−0.03)| = 0.0054717; inpercentageterms, the
error is
¸
¸
¸
¸
0.0054717
−0.0245283
¸
¸
¸
¸
×100%≈ 22.31%
f (x) = tan
_
x
4
+
π
4
_
, a = 0, x = 0.01
In Exercises 13–16, estimate y using differentials [Eq. (3)].
13. y = cosx, a =
π
6
, dx = 0.014
solution Let f (x) = cosx. Thenf

(x) = −sinx and
y ≈ dy = f

(a)dx = −sin
_
π
6
_
(0.014) = −0.007.
y = tan
2
x, a =
π
4
, dx = −0.02
15. y =
10−x
2
2+x
2
, a = 1, dx = 0.01
solution Let f (x) =
10−x
2
2+x
2
. Then
f

(x) =
(2+x
2
)(−2x) −(10−x
2
)(2x)
(2+x
2
)
2
= −
24x
(2+x
2
)
2
and
y ≈ dy = f

(a)dx = −
24
9
(0.01) = −0.026667.
y =
3−

x

x +3
, a = 1, dx = −0.1
In Exercises 17–24, estimate using the Linear Approximation and ﬁnd the error using a calculator.
17.

26−

25
solution Let f (x) =

x, a = 25, andx = 1. Thenf

(x) =
1
2
x
−1/2
andf

(a) = f

(25) =
1
10
.
• TheLinearApproximationisf ≈ f

(a)x =
1
10
(1) = 0.1.
• Theactual changeisf = f (a +x) −f (a) = f (26) −f (25) ≈ 0.0990195.
• Theerror inthisestimateis|0.0990195−0.1| = 0.000980486.
16.5
1/4
−16
1/4 19.
1

101

1
10
solution Let f (x) =
1

x
, a = 100, andx = 1. Thenf

(x) =
d
dx
(x
−1/2
) = −
1
2
x
−3/2
andf

(a) = −
1
2
(
1
1000
) =
−0.0005.
• TheLinearApproximationisf ≈ f

(a)x = −0.0005(1) = −0.0005.
• Theactual changeis
f = f (a +x) −f (a) =
1

101

1
10
= −0.000496281.
• Theerror inthisestimateis|−0.0005−(−0.000496281)| = 3.71902×10
−6
.
1

98

1
10
21. 9
1/3
−2
solution Let f (x) = x
1/3
, a = 8, andx = 1. Thenf

(x) =
1
3
x
−2/3
andf

(a) = f

(8) =
1
12
.
• TheLinearApproximationisf ≈ f

(a)x =
1
12
(1) = 0.083333.
• Theactual changeisf = f (a +x) −f (a) = f (9) −f (8) = 0.080084.
• Theerror inthisestimateis|0.080084−0.083333| ≈ 3.25×10
−3
.
June 9, 2011 LTSV SSM Second Pass
176 C HA P T E R 4 APPLICATIONS OF THE DERIVATIVE
1
(27.5)
5/3

1
243
23. sin(0.023)
solution Let f (x) = sinx, a = 0, andx = 0.023. Thenf

(x) = cosx andf

(a) = f

(0) = 1.
• TheLinearApproximationisf ≈ f

(a)x = 1· 0.023≈ 0.023.
• Theactual changeisf = f (a +x) −f (a) = f (0.023) −f (0) ≈ 0.02299797.
• Theerror inthisestimateis|0.023−0.02299797| ≈ 2×10
−6
.
tan
_
π
4
+0.01
_
−1
25. Estimatef (4.03) for f (x) asinFigure8.
(4, 2)
(10, 4)
x
y = f (x)
Tangent line
y
FIGURE 8
solution UsingtheLinear Approximation, f (4.03) ≈ f (4) + f

(4)(0.03). Fromtheﬁgure, weﬁndthat f (4) = 2
and
f

(4) =
4−2
10−4
=
1
3
.
Thus,
f (4.03) ≈ 2+
1
3
(0.03) = 2.01.
At acertainmoment, anobject inlinear motionhasvelocity 100m/s. Estimatethedistancetraveledover the
next quarter-second, andexplainhowthisisanapplicationof theLinearApproximation.
27. Whichislarger:

2.1−

2or

9.1−

9? ExplainusingtheLinearApproximation.
solution Let f (x) =

x, andx = 0.1. Thenf

(x) =
1
2
x
−1/2
andtheLinearApproximationat x = a gives
f =

a +0.1−

a ≈ f

(a)(0.1) =
1
2
a
−1/2
(0.1) =
0.05

a
Weseethat f decreasesasa increases. Inparticular

2.1−

2≈
0.05

2
islarger than

9.1−

9≈
0.05
3
Estimatesin61

−sin60

usingtheLinearApproximation. Hint: Expressθ inradians.
29. Boxofﬁcerevenueat amultiplexcinemainParisisR(p) = 3600p −10p
3
eurosper showingwhentheticket price
isp euros. CalculateR(p) for p = 9andusetheLinear ApproximationtoestimateR if p israisedor loweredby0.5
euros.
solution Let R(p) = 3600p −10p
3
. ThenR(9) = 3600(9) −10(9)
3
= 25,110euros. Moreover, R

(p) = 3600−
30p
2
, sobytheLinearApproximation,
R ≈ R

(9)p = 1170p.
If p israisedby0.5euros, thenR ≈ 585euros; ontheother hand, if p isloweredby0.5euros, thenR ≈ −585euros.
Thestopping distance for anautomobileisF(s) = 1.1s +0.054s
2
ft, wheres isthespeedinmph. UsetheLinear
Approximationtoestimatethechangeinstoppingdistanceper additional mphwhens = 35andwhens = 55.
31. Athinsilver wirehaslengthL = 18cmwhenthetemperatureisT = 30

C. EstimateL whenT decreasesto25

C
if thecoefﬁcient of thermal expansionisk = 1.9×10
−5◦
C
−1
(seeExample3).
solution Wehave
dL
dT
= kL = (1.9×10
−5
)(18) = 3.42×10
−4
cm/

C
ThechangeintemperatureisT = −5

C, sobytheLinearApproximation, thechangeinlengthisapproximately
L ≈ 3.42×10
−4
T = (3.42×10
−4
)(−5) = −0.00171cm
At T = 25

C, thelengthof thewireisapproximately17.99829cm.
Atacertainmoment, thetemperatureinasnakecagesatisﬁesdT/dt = 0.008

C/s. Estimatetheriseintemperature
over thenext 10seconds.
33. Theatmospheric pressureP at altitudeh = 20 kmis P = 5.5 kilopascals. EstimateP at altitudeh = 20.5 km
assumingthat
dP
dh
= −0.87.
solution Wehave
P ≈ P

(h)h = −0.87· 0.5≈ −0.435
June 9, 2011 LTSV SSM Second Pass
S E C T I ON 4.1 Linear Approximation and Applications 177
sothat
P(20.5) ≈ P(20) +P = 5.5−0.435≈ 5.065kilopascals.
TheresistanceR of acopper wireat temperatureT = 20

C isR = 15. Estimatetheresistanceat T = 22

C,
assumingthat dR/dT
¸
¸
T =20
= 0.06/

C.
35. Newton’s Lawof Gravitationshows that if apersonweighs w pounds onthesurfaceof theearth, thenhis or her
weight at distancex fromthecenter of theearthis
W(x) =
wR
2
x
2
(for x ≥ R)
whereR = 3960milesistheradiusof theearth(Figure9).
(a) Showthat theweight lost at altitudeh milesabovetheearth’ssurfaceisapproximatelyW ≈ −(0.0005w)h. Hint:
UsetheLinearApproximationwithdx = h.
(b) Estimatetheweight lost bya200-lbfootball player ﬂyinginajet at analtitudeof 7miles.
3
9
6
0
h
FIGURE 9 Thedistancetothecenter of theearthis3960+h miles.
solution
(a) UsingtheLinearApproximation
W ≈ W

(R)x = −
2wR
2
R
3
h = −
2wh
R
≈ −0.0005wh.
(b) Substitutew = 200andh = 7intotheresult frompart (a) toobtain
W ≈ −0.0005(200)(7) = −0.7pounds.
UsingExercise35(a), estimatethealtitudeat whicha130-lbpilot wouldweigh129.5lb.
37. A stonetossedverticallyintotheair withinitial velocityv cm/sreachesamaximumheight of h = v
2
/1960cm.
(a) Estimateh if v = 700cm/sandv = 1cm/s.
(b) Estimateh if v = 1000cm/sandv = 1cm/s.
(c) Ingeneral, doesa1cm/sincreaseinv leadtoagreater changeinh at lowor highinitial velocities? Explain.
solution A stonetossed vertically with initial velocity v cm/s attains amaximumheight of h(v) = v
2
/1960cm.
Thus, h

(v) = v/980.
(a) If v = 700andv = 1, thenh ≈ h

(v)v =
1
980
(700)(1) ≈ 0.71cm.
(b) If v = 1000andv = 1, thenh ≈ h

(v)v =
1
980
(1000)(1) = 1.02cm.
(c) Aonecentimeter per secondincreaseininitial velocityv increasesthemaximumheightbyapproximatelyv/980cm.
Accordingly, thereisabigger effect at higher velocities.
Thesides of asquarecarpet ismeasuredat 6m. Estimatethemaximumerror intheareaA of thecarpet if s is
accuratetowithin2centimeters.
In Exercises 39 and 40, use the following fact derived from Newton’s Laws: An object released at an angle θ with initial
velocity v ft/stravels a horizontal distance
s =
1
32
v
2
sin2θ ft (Figure10)
q
x
y
FIGURE 10 Trajectoryof anobject releasedat anangleθ.
39. A player located18.1ft fromthebasket launchesasuccessful jumpshot fromaheight of 10ft (level withtherimof
thebasket), at anangleθ = 34

andinitial velocityv = 25ft/s.)
(a) Showthat s ≈ 0.255θ ft for asmall changeof θ.
(b) Isit likelythat theshot wouldhavebeensuccessful if theanglehadbeenoff by2

?
June 9, 2011 LTSV SSM Second Pass
178 C HA P T E R 4 APPLICATIONS OF THE DERIVATIVE
solution Using Newton’s laws and the given initial velocity of v = 25ft/s, the shot travels s =
1
32
v
2
sin2t =
625
32
sin2t ft, wheret isinradians.
(a) If θ = 34

(i.e., t =
17
90
π), then
s ≈ s

(t )t =
625
16
cos
_
17
45
π
_
t =
625
16
cos
_
17
45
π
_
θ ·
π
180
≈ 0.255θ.
(b) If θ = 2

, thisgivess ≈ 0.51ft, inwhichcasetheshot wouldnot havebeensuccessful, havingbeenoff half a
foot.
Estimates if θ = 34

, v = 25ft/s, andv = 2.
41. Theradiusof aspherical ball ismeasuredat r = 25cm. Estimatethemaximumerror inthevolumeandsurfacearea
if r isaccuratetowithin0.5cm.
solution Thevolumeandsurfaceareaof thespherearegivenbyV =
4
3
πr
3
andS = 4πr
2
, respectively. If r = 25
andr = ±0.5, then
V ≈ V

(25)r = 4π(25)
2
(0.5) ≈ 3927cm
3
,
and
S ≈ S

(25)r = 8π(25)(0.5) ≈ 314.2cm
2
.
ThedosageD of diphenhydraminefor adogof body mass w kgis D = 4.7w
2/3
mg. Estimatethemaximum
allowableerror inw for acocker spaniel of massw = 10kgif thepercentageerror inD must belessthan3%.
43. Thevolume(in liters) and pressureP (in atmospheres) of acertain gas satisfy PV = 24. A measurement yields
V = 4withapossibleerror of ±0.3L. ComputeP andestimatethemaximumerror inthiscomputation.
solution GivenPV = 24andV = 4, itfollowsthatP = 6atmospheres. SolvingPV = 24forP yieldsP = 24V
−1
.
Thus, P

= −24V
−2
and
P ≈ P

(4)V = −24(4)
−2
(±0.3) = ±0.45atmospheres.
Inthenotationof Exercise43, assumethat ameasurement yieldsV = 4. Estimatethemaximumallowableerror
inV if P must haveanerror of lessthan0.2atm.
In Exercises 45–54, ﬁnd the linearization at x = a.
45. f (x) = x
4
, a = 1
solution Let f (x) = x
4
. Thenf

(x) = 4x
3
. Thelinearizationat a = 1is
L(x) = f

(a)(x −a) +f (a) = 4(x −1) +1= 4x −3.
f (x) =
1
x
, a = 2
47. f (θ) = sin
2
θ, a =
π
4
solution Let f (θ) = sin
2
θ. Thenf

(θ) = 2sinθ cosθ = sin2θ. Thelinearizationat a =
π
4
is
L(θ) = f

(a)(θ −a) +f (a) = 1
_
θ −
π
4
_
+
1
2
= θ −
π
4
+
1
2
.
g(x) =
x
2
x −3
, a = 4
49. y = (1+x)
−1/2
, a = 0
solution Let f (x) = (1+x)
−1/2
. Thenf

(x) = −
1
2
(1+x)
−3/2
. Thelinearizationat a = 0is
L(x) = f

(a)(x −a) +f (a) = −
1
2
x +1.
y = (1+x)
−1/2
, a = 3
51. y = (1+x
2
)
−1/2
, a = 0
solution Let f (x) = (1+ x
2
)
−1/2
. Thenf

(x) = −x(1+ x
2
)
−3/2
, f (a) = 1andf

(a) = 0, sothelinearization
at a is
L(x) = f

(a)(x −a) +f (a) = 1.
y =
1−x
1+x
, a = 4
53. y =
sinx
x
, a =
π
2
solution Let f (x) =
sinx
x
. Then
f

(x) =
x cosx −sinx
x
2
, f (a) =
2
π
, f

(a) = −
4
π
2
sothelinearizationof f (x) at a is
L(x) = f

(a)(x −a) +f (a) = −
4
π
2
_
x −
π
2
_
+
2
π
= −
4
π
2
x +
4
π
June 9, 2011 LTSV SSM Second Pass
S E C T I ON 4.1 Linear Approximation and Applications 179
y =
sinx
x
, a =
π
4
55. What isf (2) if thelinearizationof f (x) at a = 2isL(x) = 2x +4?
solution f (2) = L(2) = 2(2) +4= 8.
Computethelinearizationof f (x) = 3x − 4at a = 0anda = 2. Provemoregenerally that alinear function
coincideswithitslinearizationat x = a for all a.
57. Estimate

16.2usingthelinearizationL(x) of f (x) =

x at a = 16. Plot f (x) andL(x) onthesameset of axes
anddeterminewhether theestimateistoolargeor toosmall.
solution Letf (x) = x
1/2
, a = 16, andx = 0.2.Thenf

(x) =
1
2
x
−1/2
andf

(a) = f

(16) =
1
8
.Thelinearization
tof (x) is
L(x) = f

(a)(x −a) +f (a) =
1
8
(x −16) +4=
1
8
x +2.
Thus, wehave

16.2 ≈ L(16.2) = 4.025. Graphsof f (x) andL(x) areshownbelow. Becausethegraphof L(x) lies
abovethegraphof f (x), weexpect that theestimatefromtheLinearApproximationistoolarge.
y
x
1
2
3
4
0
5
5 10 15 25 20
f (x)
L(x)
Estimate1/

15usingasuitablelinearizationof f (x) = 1/

x. Plot f (x) andL(x) onthesameset of axes
anddeterminewhether theestimateistoolargeor toosmall. Useacalculator tocomputethepercentageerror.
In Exercises 59–67, approximate using linearization and use a calculator to compute the percentage error.
59.
1

17
solution Let f (x) = x
−1/2
, a = 16, and x = 1. Then f

(x) = −
1
2
x
−3/2
, f

(a) = f

(16) = −
1
128
and the
linearizationtof (x) is
L(x) = f

(a)(x −a) +f (a) = −
1
128
(x −16) +
1
4
= −
1
128
x +
3
8
.
Thus, wehave
1

17
≈ L(17) ≈ 0.24219. Thepercentageerror inthisestimateis
¸
¸
¸
¸
¸
¸
1

17
−0.24219
1

17
¸
¸
¸
¸
¸
¸
×100%≈ 0.14%
1
101
61.
1
(10.03)
2
solution Let f (x) = x
−2
, a = 10 and x = 0.03. Then f

(x) = −2x
−3
, f

(a) = f

(10) = −0.002 and the
linearizationtof (x) is
L(x) = f

(a)(x −a) +f (a) = −0.002(x −10) +0.01= −0.002x +0.03.
Thus, wehave
1
(10.03)
2
≈ L(10.03) = −0.002(10.03) +0.03= 0.00994.
Thepercentageerror inthisestimateis
¸
¸
¸
¸
¸
¸
1
(10.03)
2
−0.00994
1
(10.03)
2
¸
¸
¸
¸
¸
¸
×100%≈ 0.0027%
(17)
1/4
63. (64.1)
1/3
solution Letf (x) = x
1/3
,a = 64,andx = 0.1.Thenf

(x) =
1
3
x
−2/3
,f

(a) = f

(64) =
1
48
andthelinearization
tof (x) is
L(x) = f

(a)(x −a) +f (a) =
1
48
(x −64) +4=
1
48
x +
8
3
.
Thus, wehave(64.1)
1/3
≈ L(64.1) ≈ 4.002083. Thepercentageerror inthisestimateis
¸
¸
¸
¸
¸
(64.1)
1/3
−4.002083
(64.1)
1/3
¸
¸
¸
¸
¸
×100%≈ 0.000019%
June 9, 2011 LTSV SSM Second Pass
180 C HA P T E R 4 APPLICATIONS OF THE DERIVATIVE
(1.2)
5/3
65. tan(0.04)
solution Let f (x) = tanx and a = 0. Then f

(x) = sec
2
x, f (a) = tan0 = 0, and f

(a) = sec
2
0 = 1. The
linearizationof f (x) isthen
L(x) = f

(a)(x −a) +f (a) = 1(x −0) +0= x
Thus, wehavetan(0.04) ≈ 0.04. Thepercentageerror inthisestimateis
¸
¸
¸
¸
tan(0.04) −0.04
tan(0.04)
¸
¸
¸
¸
×100%≈ 0.053%.
cos
_
3.1
4
_
67.
(3.1)/2
sin(3.1/2)
solution Let f (x) =
x
sinx
anda =
π
2
. Then
f

(x) =
sinx −x cosx
sin
2
x
, f (a) =
π
2
, f

(a) = 1
Thelinearizationof f (x) isthen
L(x) = f

(a)(x −a) +f (a) = x −
π
2
+
π
2
= x
Thus, wehave
(3.1/2)
sin(3.1/2)
≈ L
_
3.1
2
_
= 1.55. Thepercentageerror inthisestimateis
¸
¸
¸
¸
¸
¸
(3.1/2)
sin(3.1/2)
−1.55
(3.1/2)
sin(3.1/2)
¸
¸
¸
¸
¸
¸
×100%≈ 0.022%
ComputethelinearizationL(x) of f (x) = x
2
−x
3/2
at a = 4. Thenplot f (x) −L(x) andﬁndaninterval I
arounda = 4suchthat |f (x) −L(x)| ≤ 0.1for x ∈ I.
69. Showthat theLinear Approximationtof (x) =

x at x = 9yieldstheestimate

9+h − 3≈
1
6
h. Set K = 0.01
and showthat |f

(x)| ≤ K for x ≥ 9. Then verify numerically that theerror E satisﬁes Eq. (5) for h = 10
−n
, for
1≤ n ≤ 4.
solution Letf (x) =

x. Thenf (9) = 3, f

(x) =
1
2
x
−1/2
andf

(9) =
1
6
. Therefore, bytheLinearApproximation,
f (9+h) −f (9) =

9+h −3≈
1
6
h.
Moreover, f

(x) = −
1
4
x
−3/2
, so|f

(x)| =
1
4
x
−3/2
. Becausethisisadecreasingfunction, it followsthat for x ≥ 9,
K = max|f

(x)| ≤ |f

(9)| =
1
108
< 0.01.
Fromthefollowingtable, weseethat for h = 10
−n
, 1≤ n ≤ 4, E ≤
1
2
Kh
2
.
h E = |

9+h −3−
1
6
h|
1
2
Kh
2
10
−1
4.604×10
−5
5.00×10
−5
10
−2
4.627×10
−7
5.00×10
−7
10
−3
4.629×10
−9
5.00×10
−9
10
−4
4.627×10
−11
5.00×10
−11
TheLinear Approximationtof (x) = tanx at x =
π
4
yieldstheestimatetan
_
π
4
+h
_
−1≈ 2h. Set K = 6.2
andshow, usingaplot, that |f

(x)| ≤ K for x ∈ [
π
4
,
π
4
+ 0.1]. Thenverify numerically that theerror E satisﬁes
Eq. (5) for h = 10
−n
, for 1≤ n ≤ 4.
Further Insights and Challenges
71. Compute dy/dx at the point P = (2, 1) on the curve y
3
+ 3xy = 7 and show that the linearization at P is
L(x) = −
1
3
x +
5
3
. UseL(x) toestimatethey-coordinateof thepoint onthecurvewherex = 2.1.
solution Differentiatingbothsidesof theequationy
3
+3xy = 7withrespect tox yields
3y
2
dy
dx
+3x
dy
dx
+3y = 0,
so
dy
dx
= −
y
y
2
+x
.
June 9, 2011 LTSV SSM Second Pass
S E C T I ON 4.2 Extreme Values 181
Thus,
dy
dx
¸
¸
¸
¸
(2,1)
= −
1
1
2
+2
= −
1
3
,
andthelinearizationat P = (2, 1) is
L(x) = 1−
1
3
(x −2) = −
1
3
x +
5
3
.
Finally, whenx = 2.1, weestimatethat they-coordinateof thepoint onthecurveis
y ≈ L(2.1) = −
1
3
(2.1) +
5
3
= 0.967.
Apply themethodof Exercise71toP = (0.5, 1) ony
5
+ y − 2x = 1toestimatethey-coordinateof thepoint
onthecurvewherex = 0.55.
73. Apply themethodof Exercise71toP = (−1, 2) ony
4
+ 7xy = 2toestimatethesolutionof y
4
− 7.7y = 2near
y = 2.
solution Differentiatingbothsidesof theequationy
4
+7xy = 2withrespect tox yields
4y
3
dy
dx
+7x
dy
dx
+7y = 0,
so
dy
dx
= −
7y
4y
3
+7x
.
Thus,
dy
dx
¸
¸
¸
¸
(−1,2)
= −
7(2)
4(2)
3
+7(−1)
= −
14
25
,
andthelinearizationat P = (−1, 2) is
L(x) = 2−
14
25
(x +1) = −
14
25
x +
36
25
.
Finally, theequationy
4
−7.7y = 2correspondstox = −1.1, soweestimatethesolutionof thisequationnear y = 2is
y ≈ L(−1.1) = −
14
25
(−1.1) +
36
25
= 2.056.
Showthat for anyreal number k, (1+x)
k
≈ 1+kx for small x. Estimate(1.02)
0.7
and(1.02)
−0.3
.
75. Let f = f (5+h) −f (5), wheref (x) = x
2
. Verifydirectlythat E = |f −f

(5)h| satisﬁes(5) withK = 2.
solution Let f (x) = x
2
. Then
f = f (5+h) −f (5) = (5+h)
2
−5
2
= h
2
+10h
and
E = |f −f

(5)h| = |h
2
+10h −10h| = h
2
=
1
2
(2)h
2
=
1
2
Kh
2
.
Let f = f (1+h) −f (1) wheref (x) = x
−1
. Showdirectlythat E = |f −f

(1)h| isequal toh
2
/(1+h).
Thenprovethat E ≤ 2h
2
if −
1
2
≤ h ≤
1
2
. Hint: Inthiscase,
1
2
≤ 1+h ≤
3
2
.
4.2 Extreme Values
Preliminary Questions
1. What isthedeﬁnitionof acritical point?
solution Acritical pointisavalueof theindependentvariablex inthedomainof afunctionf atwhicheitherf

(x) = 0
or f

(x) doesnot exist.
In Questions 2 and 3, choose the correct conclusion.
2. If f (x) isnot continuouson[0, 1], then
(a) f (x) hasnoextremevalueson[0, 1].
(b) f (x) might not haveanyextremevalueson[0, 1].
solution Thecorrect responseis (b): f (x) might not haveany extremevalues on[0, 1]. Although[0, 1] is closed,
becausef isnot continuous, thefunctionisnot guaranteedtohaveanyextremevalueson[0, 1].
June 9, 2011 LTSV SSM Second Pass
182 C HA P T E R 4 APPLICATIONS OF THE DERIVATIVE
3. If f (x) iscontinuousbut hasnocritical pointsin[0, 1], then
(a) f (x) hasnominor maxon[0, 1].
(b) Either f (0) or f (1) istheminimumvalueon[0, 1].
solution Thecorrect responseis (b): either f (0) or f (1) is theminimumvalueon[0, 1]. Remember that extreme
valuesoccur either at critical pointsor endpoints. If acontinuousfunctiononaclosedinterval hasnocritical points, the
extremevaluesmust occur at theendpoints.
4. Fermat’s Theoremdoes not claimthat if f

(c) = 0, thenf (c) is alocal extremevalue(this is false). What does
Fermat’sTheoremassert?
solution Fermat’sTheoremclaims: If f (c) isalocal extremevalue, theneither f

(c) = 0or f

(c) doesnot exist.
Exercises
1. Thefollowingquestionsrefer toFigure15.
(a) Howmanycritical pointsdoesf (x) haveon[0, 8]?
(b) What isthemaximumvalueof f (x) on[0, 8]?
(c) What arethelocal maximumvaluesof f (x)?
(d) Findaclosedinterval onwhichboththeminimumandmaximumvaluesof f (x) occur at critical points.
(e) Findaninterval onwhichtheminimumvalueoccursat anendpoint.
8 3 4 5 6 7 2 1
2
3
4
5
6
1
f (x)
x
y
FIGURE 15
solution
(a) f (x) hasthreecritical pointsontheinterval [0, 8]: at x = 3, x = 5andx = 7. Twoof these, x = 3andx = 5, are
wherethederivativeiszeroandone, x = 7, iswherethederivativedoesnot exist.
(b) Themaximumvalueof f (x) on[0, 8] is6; thefunctiontakesthisvalueat x = 0.
(c) f (x) achievesalocal maximumof 5at x = 5.
(d) Answersmayvary. Oneexampleistheinterval [4, 8]. Another is[2, 6].
(e) Answersmayvary. Theeasiestwaytoensurethisistochooseaninterval onwhichthegraphtakesnolocal minimum.
Oneexampleis[0, 2].
Statewhether f (x) = x
−1
(Figure16) hasaminimumor maximumvalueonthefollowingintervals:
(a) (0, 2) (b) (1, 2) (c) [1, 2]
In Exercises 3–20, ﬁnd all critical points of the function.
3. f (x) = x
2
−2x +4
solution Let f (x) = x
2
−2x +4. Thenf

(x) = 2x −2= 0impliesthat x = 1isthelonecritical point of f .
f (x) = 7x −2
5. f (x) = x
3

9
2
x
2
−54x +2
solution Let f (x) = x
3

9
2
x
2
− 54x + 2. Then f

(x) = 3x
2
− 9x − 54 = 3(x + 3)(x − 6) = 0implies that
x = −3andx = 6arethecritical pointsof f .
f (t ) = 8t
3
−t
2
7. f (x) = x
−1
−x
−2
solution Let f (x) = x
−1
−x
−2
. Then
f

(x) = −x
−2
+2x
−3
=
2−x
x
3
= 0
impliesthat x = 2istheonly critical point of f . Thoughf

(x) doesnot exist at x = 0, thisisnot acritical point of f
becausex = 0isnot inthedomainof f .
g(z) =
1
z −1

1
z
9. f (x) =
x
x
2
+1
solution Let f (x) =
x
x
2
+1
. Thenf

(x) =
1−x
2
(x
2
+1)
2
= 0impliesthat x = ±1arethecritical pointsof f .
f (x) =
x
2
x
2
−4x +8
June 9, 2011 LTSV SSM Second Pass
S E C T I ON 4.2 Extreme Values 183
11. f (t ) = t −4

t +1
solution Let f (t ) = t −4

t +1. Then
f

(t ) = 1−
2

t +1
= 0
impliesthat t = 3isacritical point of f . Becausef

(t ) doesnot exist at t = −1, thisisanother critical point of f .
f (t ) = 4t −
_
t
2
+1
13. f (x) = x
2
_
1−x
2
solution Let f (x) = x
2
_
1−x
2
. Then
f

(x) = −
x
3
_
1−x
2
+2x
_
1−x
2
=
2x −3x
3
_
1−x
2
.
This derivativeis 0whenx = 0andwhenx = ±

2/3; thederivativedoes not exist whenx = ±1. All ﬁveof these
valuesarecritical pointsof f
f (x) = x +|2x +1|
15. g(θ) = sin
2
θ
solution Let g(θ) = sin
2
θ. Theng

(θ) = 2sinθ cosθ = sin2θ = 0impliesthat
θ =

2
isacritical valueof g for all integer valuesof n.
R(θ) = cosθ +sin
2
θ
17. Let f (x) = x
2
−4x +1.
(a) Findthecritical point c of f (x) andcomputef (c).
(b) Computethevalueof f (x) at theendpointsof theinterval [0, 4].
(c) Determinetheminandmaxof f (x) on[0, 4].
(d) Findtheextremevaluesof f (x) on[0, 1].
solution Let f (x) = x
2
−4x +1.
(a) Thenf

(c) = 2c −4= 0impliesthat c = 2isthesolecritical point of f . Wehavef (2) = −3.
(b) f (0) = f (4) = 1.
(c) Usingtheresultsfrom(a) and(b), weﬁndthemaximumvalueof f on[0, 4] is1andtheminimumvalueis−3.
(d) Wehavef (1) = −2. Hencethemaximumvalueof f on[0, 1] is1andtheminimumvalueis−2.
Findtheextremevaluesof f (x) = 2x
3
−9x
2
+12x on[0, 3] and[0, 2].
19. Findthecritical pointsof f (x) = sinx +cosx anddeterminetheextremevalueson
_
0,
π
2
_
.
solution
• Let f (x) = sinx + cosx. Thenontheinterval
_
0,
π
2
_
, wehavef

(x) = cosx − sinx = 0at x =
π
4
, theonly
critical point of f inthisinterval.
• Sincef (
π
4
) =

2andf (0) = f (
π
2
) = 1, themaximumvalueof f on
_
0,
π
2
_
is

2, whiletheminimumvalue
is1.
Computethecritical pointsof h(t ) = (t
2
−1)
1/3
. Checkthatyour answer isconsistentwithFigure17. Thenﬁnd
theextremevaluesof h(t ) on[0, 1] and[0, 2].
21. Plot f (x) = 2

x − x on [0, 4] and determinethemaximumvaluegraphically. Then verify your answer
usingcalculus.
solution Thegraphof y = 2

x − x over theinterval [0, 4] isshownbelow. Fromthegraph, weseethat at x = 1,
thefunctionachievesitsmaximumvalueof 1.
y
x
0.2
0.4
0.6
0.8
0
1
1 2 3 4
To verify theinformationobtainedfromtheplot, let f (x) = 2

x − x. Thenf

(x) = x
−1/2
− 1. Solvingf

(x) = 0
yieldsthecritical pointsx = 0andx = 1. Becausef (0) = f (4) = 0andf (1) = 1, weseethat themaximumvalueof
f on[0, 4] is1.
Plot f (x) = 2x
3
− 9x
2
+ 12x on[0, 3] andlocatetheextremevalues graphically. Thenverify your answer
usingcalculus.
June 9, 2011 LTSV SSM Second Pass
184 C HA P T E R 4 APPLICATIONS OF THE DERIVATIVE
23. Approximatethecritical pointsof g(x) = x cosx onI = [0, 2π] andestimatetheminimumvalueof g(x)
onI.
solution g

(x) = cosx − x sinx, so g

(x) = 0whenx ≈ 0.860334andwhenx ≈ 3.425618. Evaluatingg at the
endpointsaswell asat thesecritical pointsgivesg(0) = 0, g(0.860334) ≈ 0.561096, g(3.425618) ≈ −3.288371, and
g(2π) = 2π ≈ 6.28. Hencetheminimumvalueof g(x) onI is≈ −3.288371at x ≈ 3.425618.
Approximatethecritical pointsof g(x) = tan
2
x −5x onI =
_

π
2
,
π
2
_
andestimatetheminimumvalueof
g(x) onI.
In Exercises 25–58, ﬁnd the min and max of the function on the given interval by comparing values at the critical points
and endpoints.
25. y = 2x
2
+4x +5, [−2, 2]
solution Let f (x) = 2x
2
+ 4x + 5. Thenf

(x) = 4x + 4= 0impliesthat x = −1istheonly critical point of f .
Theminimumof f ontheinterval [−2, 2] isf (−1) = 3, whereasitsmaximumisf (2) = 21. (Note: f (−2) = 5.)
y = 2x
2
+4x +5, [0, 2]
27. y = 6t −t
2
, [0, 5]
solution Letf (t ) = 6t −t
2
. Thenf

(t ) = 6−2t = 0impliesthatt = 3istheonlycritical pointof f . Theminimum
of f ontheinterval [0, 5] isf (0) = 0, whereasthemaximumisf (3) = 9. (Note: f (5) = 5.)
y = 6t −t
2
, [4, 6]
29. y = x
3
−6x
2
+8, [1, 6]
solution Let f (x) = x
3
− 6x
2
+ 8. Thenf

(x) = 3x
2
− 12x = 3x(x − 4) = 0impliesthat x = 0andx = 4are
thecritical points of f . Theminimumof f ontheinterval [1, 6] is f (4) = −24, whereas themaximumis f (6) = 8.
(Note: f (1) = 3andthecritical point x = 0isnot intheinterval [1, 6].)
y = x
3
+x
2
−x, [−2, 2]
31. y = 2t
3
+3t
2
, [1, 2]
solution Let f (t ) = 2t
3
+ 3t
2
. Thenf

(t ) = 6t
2
+ 6t = 6t (t + 1) = 0implies that t = 0andt = −1arethe
critical pointsof f . Theminimumof f ontheinterval [1, 2] isf (1) = 5, whereasthemaximumisf (2) = 28. (Note:
Neither critical pointsareintheinterval [1, 2].)
y = x
3
−12x
2
+21x, [0, 2]
33. y = z
5
−80z, [−3, 3]
solution Let f (z) = z
5
− 80z. Then f

(z) = 5z
4
− 80 = 5(z
4
− 16) = 5(z
2
+ 4)(z + 2)(z − 2) = 0 implies
that z = ±2arethecritical pointsof f . Theminimumvalueof f ontheinterval [−3, 3] isf (2) = −128, whereasthe
maximumisf (−2) = 128. (Note: f (−3) = 3andf (3) = −3.)
y = 2x
5
+5x
2
, [−2, 2] 35. y =
x
2
+1
x −4
, [5, 6]
solution Let f (x) =
x
2
+1
x −4
. Then
f

(x) =
(x −4) · 2x −(x
2
+1) · 1
(x −4)
2
=
x
2
−8x −1
(x −4)
2
= 0
impliesx = 4±

17arecritical pointsof f . x = 4isnot acritical point becausex = 4isnot inthedomainof f . On
theinterval [5, 6], theminimumof f isf (6) =
37
2
= 18.5, whereasthemaximumof f isf (5) = 26. (Note: Thecritical
pointsx = 4±

17arenot intheinterval [5, 6].)
y =
1−x
x
2
+3x
, [1, 4]
37. y = x −
4x
x +1
, [0, 3]
solution Let f (x) = x −
4x
x +1
. Then
f

(x) = 1−
4
(x +1)
2
=
(x −1)(x +3)
(x +1)
2
= 0
impliesthatx = 1andx = −3arecritical pointsof f . x = −1isnotacritical pointbecausex = −1isnotinthedomain
of f . Theminimumof f ontheinterval [0, 3] is f (1) = −1, whereas themaximumis f (0) = f (3) = 0. (Note: The
critical point x = −3isnot intheinterval [0, 3].)
y = 2
_
x
2
+1−x, [0, 2]
39. y = (2+x)
_
2+(2−x)
2
, [0, 2]
solution Let f (x) = (2+x)
_
2+(2−x)
2
. Then
f

(x) =
_
2+(2−x)
2
−(2+x)(2+(2−x)
2
)
−1/2
(2−x) =
2(x −1)
2
_
2+(2−x)
2
= 0
impliesthat x = 1isthecritical point of f . Ontheinterval [0, 2], theminimumisf (0) = 2

6≈ 4.9andthemaximum
isf (2) = 4

2≈ 5.66. (Note: f (1) = 3

3≈ 5.2.)
y =
_
1+x
2
−2x, [0, 1]
June 9, 2011 LTSV SSM Second Pass
S E C T I ON 4.2 Extreme Values 185
41. y =
_
x +x
2
−2

x, [0, 4]
solution Let f (x) =
_
x +x
2
−2

x. Then
f

(x) =
1
2
(x +x
2
)
−1/2
(1+2x) −x
−1/2
=
1+2x −2

1+x
2

x

1+x
= 0
implies that x = 0andx =

3
2
arethecritical points of f . Neither x = −1nor x = −

3
2
is acritical point because
neither isinthedomainof f . Ontheinterval [0, 4], theminimumof f isf
_√
3
2
_
≈ −0.589980andthemaximumis
f (4) ≈ 0.472136. (Note: f (0) = 0.)
y = (t −t
2
)
1/3
, [−1, 2]
43. y = sinx cosx,
_
0,
π
2
_
solution Letf (x) = sinx cosx =
1
2
sin2x. Ontheinterval
_
0,
π
2
_
, f

(x) = cos2x = 0whenx =
π
4
. Theminimum
of f onthisinterval isf (0) = f (
π
2
) = 0, whereasthemaximumisf (
π
4
) =
1
2
.
y = x +sinx, [0, 2π]
45. y =

2θ −secθ,
_
0,
π
3
_
solution Let f (θ) =

2θ − secθ. Ontheinterval [0,
π
3
], f

(θ) =

2− secθ tanθ = 0at θ =
π
4
. Theminimum
valueof f on this interval is f (0) = −1, whereas themaximumvalueover this interval is f (
π
4
) =

2(
π
4
− 1) ≈
−0.303493. (Note: f (
π
3
) =

2
π
3
−2≈ −0.519039.)
y = cosθ +sinθ, [0, 2π]
47. y = θ −2sinθ, [0, 2π]
solution Let g(θ) = θ − 2sinθ. On theinterval [0, 2π], g

(θ) = 1− 2cosθ = 0 at θ =
π
3
and θ =
5
3
π. The
minimumof g onthisinterval isg(
π
3
) =
π
3

3≈ −0.685andthemaximumisg(
5
3
π) =
5
3
π +

3≈ 6.968. (Note:
g(0) = 0andg(2π) = 2π ≈ 6.283.)
y = 4sin
3
θ −3cos
2
θ, [0, 2π]
49. y = tanx −2x, [0, 1]
solution Let f (x) = tanx −2x. Thenontheinterval [0, 1], f

(x) = sec
2
x −2= 0at x =
π
4
. Theminimumof f
isf (
π
4
) = 1−
π
2
≈ −0.570796andthemaximumisf (0) = 0. (Note: f (1) = tan1−2≈ −0.442592.)
y = sec
2
x −2tanx,
_

π
6
,
π
3
_ 51. Let f (θ) = 2sin2θ +sin4θ.
(a) Showthat θ isacritical point if cos4θ = −cos2θ.
(b) Show, usingaunit circle, that cosθ
1
= −cosθ
2
if andonlyif θ
1
= π ±θ
2
+2πk for aninteger k.
(c) Showthat cos4θ = −cos2θ if andonlyif θ =
π
2
+πk or θ =
π
6
+
_
π
3
_
k.
(d) Findthesixcritical pointsof f (θ) on[0, 2π] andﬁndtheextremevaluesof f (θ) onthisinterval.
(e) Checkyour resultsagainst agraphof f (θ).
solution f (θ) = 2sin2θ + sin4θ isdifferentiableat all θ, sotheway toﬁndthecritical pointsistoﬁndall points
suchthat f

(θ) = 0.
(a) f

(θ) = 4cos2θ +4cos4θ. If f

(θ) = 0, then4cos4θ = −4cos2θ, socos4θ = −cos2θ.
(b) Giventhepoint(cosθ, sinθ) atangleθ ontheunitcircle, therearetwopointswithx coordinate−cosθ. Thegraphic
showsthesetwopoints, whichare:
• Thepoint (cos(θ +π), sin(θ +π)) ontheoppositeendof theunit circle.
• Thepoint (cos(π −θ), sin(θ −π)) obtainedbyreﬂectingthroughthey axis.
If we include all angles representing these points on the circle, we ﬁnd that cosθ
1
= −cosθ
2
if and only if θ
1
=
(π +θ
2
) +2πk or θ
1
= (π −θ
2
) +2πk for integersk.
(c) Using(b), werecognizethat cos4θ = −cos2θ if 4θ = 2θ + π + 2πk or 4θ = π − 2θ + 2πk. Solvingfor θ, we
obtainθ =
π
2
+kπ or θ =
π
6
+
π
3
k.
(d) Toﬁndall θ, 0≤ θ < 2π indicatedby(c), weusethefollowingtable:
k 0 1 2 3 4 5
π
2
+kπ
π
2

2
π
6
+
π
3
k
π
6
π
2

6

6

2
11π
6
June 9, 2011 LTSV SSM Second Pass
186 C HA P T E R 4 APPLICATIONS OF THE DERIVATIVE
Thecritical pointsintherange[0, 2π] are
π
6
,
π
2
,

6
,

6
,

2
, and
11π
6
. Onthisinterval, themaximumvalueisf (
π
6
) =
f (

6
) =
3

3
2
andtheminimumvalueisf (

6
) = f (
11π
6
) = −
3

3
2
.
(e) Thegraphof f (θ) = 2sin2θ +sin4θ isshownhere:
x
1
2 3
4
5
6
1
2
−1
−2
y
Wecanseethat therearesixﬂat pointsonthegraphbetween0and2π, aspredicted. Thereare4local extrema, andtwo
pointsat (
π
2
, 0) and(

2
, 0) wherethegraphhasneither alocal maximumnor alocal minimum.
Findthecritical pointsof f (x) = 2cos3x +3cos2x in[0, 2π]. Checkyour answer against agraphof f (x).
In Exercises 53–56, ﬁnd the critical points and the extreme values on [0, 4]. In Exercises 55 and 56, refer to Figure 18.
y = |x
2
+ 4x − 12|
2 −6
10
20
30
y = |cos x|
1
π
2
π 3π
2

π
2
x x
y y
FIGURE 18
53. y = |x −2|
solution Let f (x) = |x −2|. For x < 2, wehavef

(x) = −1. For x > 2, wehavef

(x) = 1. Nowasx →2−, we
have
f (x) −f (2)
x −2
=
(2−x) −0
x −2
→−1; whereasasx →2+, wehave
f (x) −f (2)
x −2
=
(x −2) −0
x −2
→1. Therefore,
f

(2) = lim
x→2
f (x) −f (2)
x −2
doesnot exist andthelonecritical point of f isx = 2. Alternately, weexaminethegraphof
f (x) = |x −2| shownbelow.
Toﬁndtheextremum, wecheck thevaluesof f (x) at thecritical point andtheendpoints. f (0) = 2, f (4) = 2, and
f (2) = 0. f (x) takesitsminimumvalueof 0at x = 2, anditsmaximumof 2at x = 0andat x = 4.
y
x
0.5
1
1.5
2
0 3 4 2 1
y = |3x −9|
55. y = |x
2
+4x −12|
solution Let f (x) = |x
2
+4x −12| = |(x +6)(x −2)|. Fromthegraphof f inFigure18, weseethat f

(x) does
not exist at x = −6andat x = 2, so thesearecritical points of f . Thereis also acritical point betweenx = −6and
x = 2at whichf

(x) = 0. For −6 < x < 2, f (x) = −x
2
− 4x + 12, so f

(x) = −2x − 4 = 0whenx = −2. On
theinterval [0, 4] theminimumvalueof f isf (2) = 0andthemaximumvalueisf (4) = 20. (Note: f (0) = 12andthe
critical pointsx = −6andx = −2arenot intheinterval.)
y = | cosx|
In Exercises 57–60, verify Rolle’s Theorem for the given interval.
57. f (x) = x +x
−1
,
_
1
2
, 2
_
solution Becausef iscontinuouson[
1
2
, 2], differentiableon(
1
2
, 2) and
f
_
1
2
_
=
1
2
+
1
1
2
=
5
2
= 2+
1
2
= f (2),
wemayconcludefromRolle’sTheoremthatthereexistsac ∈ (
1
2
, 2) atwhichf

(c) = 0.Here,f

(x) = 1−x
−2
=
x
2
−1
x
2
,
sowemaytakec = 1.
f (x) = sinx,
_
π
4
,

4
_
June 9, 2011 LTSV SSM Second Pass
S E C T I ON 4.2 Extreme Values 187
59. f (x) =
x
2
8x −15
, [3, 5]
solution Becausef is continuous on[3, 5], differentiableon(3, 5) andf (3) = f (5) = 1, wemay concludefrom
Rolle’sTheoremthat thereexistsac ∈ (3, 5) at whichf

(c) = 0. Here,
f

(x) =
(8x −15)(2x) −8x
2
(8x −15)
2
=
2x(4x −15)
(8x −15)
2
,
sowemaytakec =
15
4
.
f (x) = sin
2
x −cos
2
x,
_
π
4
,

4
_ 61. Provethat f (x) = x
5
+2x
3
+4x −12haspreciselyonereal root.
solution Let’sﬁrst establishthef (x) = x
5
+ 2x
3
+ 4x − 12hasat least oneroot. Becausef isapolynomial, it is
continuousfor all x. Moreover, f (0) = −12 < 0andf (2) = 44 > 0. Therefore, by theIntermediateValueTheorem,
thereexistsac ∈ (0, 2) suchthat f (c) = 0.
Next, weprovethatthisistheonlyroot. Wewill useproof bycontradiction. Supposef (x) = x
5
+2x
3
+4x −12has
tworeal roots, x = a andx = b. Thenf (a) = f (b) = 0andRolle’sTheoremguaranteesthat thereexistsac ∈ (a, b) at
whichf

(c) = 0. However, f

(x) = 5x
4
+6x
2
+4≥ 4for all x, sothereisnoc ∈ (a, b) atwhichf

(c) = 0. Basedon
thiscontradiction, weconcludethat f (x) = x
5
+ 2x
3
+ 4x − 12cannot havemorethanonereal root. Finally, f must
havepreciselyonereal root.
Provethat f (x) = x
3
+3x
2
+6x haspreciselyonereal root.
63. Provethat f (x) = x
4
+5x
3
+4x hasnoroot c satisfyingc > 0. Hint: Notethat x = 0isaroot andapplyRolle’s
Theorem.
solution We will proceed by contradiction. Note that f (0) = 0 and suppose that there exists a c > 0 such that
f (c) = 0. Thenf (0) = f (c) = 0andRolle’sTheoremguarantees that thereexists ad ∈ (0, c) suchthat f

(d) = 0.
However, f

(x) = 4x
3
+ 15x
2
+ 4 > 4for all x > 0, so thereis no d ∈ (0, c) suchthat f

(d) = 0. Basedonthis
contradiction, weconcludethat f (x) = x
4
+5x
3
+4x hasnoroot c satisfyingc > 0.
Provethat c = 4isthelargest root of f (x) = x
4
−8x
2
−128.
65. Thepositionof amass oscillatingat theendof aspringis s(t ) = Asinωt , whereA is theamplitudeandω is the
angular frequency. Showthat thespeed|v(t )| isat amaximumwhentheaccelerationa(t ) iszeroandthat |a(t )| isat a
maximumwhenv(t ) iszero.
solution Let s(t ) = Asinωt . Then
v(t ) =
ds
dt
= Aωcosωt
and
a(t ) =
dv
dt
= −Aω
2
sinωt.
Thus, thespeed
|v(t )| = |Aωcosωt |
isamaximumwhen| cosωt | = 1, whichispreciselywhensinωt = 0; thatis, thespeed|v(t )| isatamaximumwhenthe
accelerationa(t ) iszero. Similarly,
|a(t )| = |Aω
2
sinωt |
isamaximumwhen| sinωt | = 1, whichispreciselywhencosωt = 0; that is, |a(t )| isat amaximumwhenv(t ) iszero.
TheconcentrationC(t ) (inmg/cm
3
) of adruginapatient’sbloodstreamafter t hoursis
C(t ) =
0.016t
t
2
+4t +4
Findthemaximumconcentrationinthetimeinterval [0, 8] andthetimeat whichit occurs.
67. In1919, physicistAlfredBetzarguedthat themaximumefﬁciencyof awindturbineisaround59%. If windentersa
turbinewithspeedv
1
andexitswithspeedv
2
, thenthepower extractedisthedifferenceinkineticenergyper unit time:
P =
1
2
mv
2
1

1
2
mv
2
2
watts
wheremisthemassof windﬂowingthroughtherotor per unittime(Figure19). Betzassumedthatm = ρA(v
1
+v
2
)/2,
whereρ isthedensity of air andA istheareaswept out by therotor. Windﬂowingundisturbedthroughthesamearea
A would have mass per unit time ρAv
1
and power P
0
=
1
2
ρAv
3
1
. The fraction of power extracted by the turbine is
F = P/P
0
.
(a) Showthat F dependsonlyontheratior = v
2
/v
1
andisequal toF(r) =
1
2
(1−r
2
)(1+r), where0≤ r ≤ 1.
(b) Showthat themaximumvalueof F(r), calledtheBetzLimit, is16/27≈ 0.59.
(c) Explainwhy Betz’s formulafor F(r) is not meaningful for r closeto zero. Hint: Howmuchwindwould
passthroughtheturbineif v
2
werezero? Isthisrealistic?
June 9, 2011 LTSV SSM Second Pass
188 C HA P T E R 4 APPLICATIONS OF THE DERIVATIVE
1
0.1
0.2
0.3
0.5
0.4
0.6
0.5
r
F
(A) Wind flowing through aturbine. (B) F is thefraction of energy
extracted by theturbineas a
function of r = v
2
/v
1.
v
1
v
2
FIGURE 19
solution
(a) Wenotethat
F =
P
P
0
=
1
2
ρA(v
1
+v
2
)
2
(v
2
1
−v
2
2
)
1
2
ρAv
3
1
=
1
2
v
2
1
−v
2
2
v
2
1
·
v
1
+v
2
v
1
=
1
2
_
1−
v
2
2
v
2
1
_
_
1+
v
2
v
1
_
=
1
2
(1−r
2
)(1+r).
(b) Basedonpart (a),
F

(r) =
1
2
(1−r
2
) −r(1+r) = −
3
2
r
2
−r +
1
2
.
Therootsof thisquadraticarer = −1andr =
1
3
. Now, F(0) =
1
2
, F(1) = 0and
F
_
1
3
_
=
1
2
·
8
9
·
4
3
=
16
27
≈ 0.59.
Thus, theBetz Limit is16/27≈ 0.59.
(c) If v
2
were0, thennoair wouldbepassingthroughtheturbine, whichisnot realistic.
TheBohr radiusa
0
of thehydrogenatomisthevalueof r that minimizestheenergy
E(r) =
¯
h
2
2mr
2

e
2

0
r
where
¯
h, m, e, and
0
arephysical constants. Showthat a
0
= 4π

h
2
/(me
2
). Assumethat theminimumoccursat a
critical point, assuggestedbyFigure20.
69. Theresponseof acircuitorotheroscillatorysystemtoaninputof frequencyω (“omega”) isdescribedbythefunction
φ(ω) =
1
_

2
0
−ω
2
)
2
+4D
2
ω
2
Bothω
0
(thenatural frequencyof thesystem) andD(thedampingfactor) arepositiveconstants. Thegraphof φ iscalleda
resonancecurve, andthepositivefrequencyω
r
> 0, whereφ takesitsmaximumvalue, if itexists, iscalledtheresonant
frequency. Showthatω
r
=
_
ω
2
0
−2D
2
if 0< D < ω
0
/

2andthatnoresonantfrequencyexistsotherwise(Figure21).
w
(A) D = 0.01 (B) D = 0.2
2 2 w
r
(C) D = 0.75 (no resonance)
50
f f f
w
2 w
r
1
2
3
w
3 1 2
1
0.5
FIGURE 21 Resonancecurveswithω
0
= 1.
solution Let φ(ω) = ((ω
2
0
−ω
2
)
2
+4D
2
ω
2
)
−1/2
. Then
φ

(ω) =
2ω((ω
2
0
−ω
2
) −2D
2
)
((ω
2
0
−ω
2
)
2
+4D
2
ω
2
)
3/2
andthenon-negativecritical points areω = 0andω =
_
ω
2
0
−2D
2
. Thelatter critical point is positiveif andonly if
ω
2
0
−2D
2
> 0, andsincewearegivenD > 0, thisisequivalent to0< D < ω
0
/

2.
June 9, 2011 LTSV SSM Second Pass
S E C T I ON 4.2 Extreme Values 189
Deﬁneω
r
=
_
ω
2
0
−2D
2
. Now, φ(0) = 1/ω
2
0
andφ(ω) →0asω →∞. Finally,
φ(ω
r
) =
1
2D
_
ω
2
0
−D
2
,
which, for 0< D < ω
0
/

2, islarger than1/ω
2
0
. Hence, thepoint ω =
_
ω
2
0
−2D
2
, if deﬁned, isalocal maximum.
Beesbuildhoneycombstructuresoutof cellswithahexagonal baseandthreerhombus-shapedfacesontop, asin
Figure22. Wecanshowthat thesurfaceareaof thiscell is
A(θ) = 6hs +
3
2
s
2
(

3 cscθ −cotθ)
withh, s, andθ asindicatedintheﬁgure. Remarkably, bees“know” whichangleθ minimizesthesurfacearea(and
thereforerequirestheleast amount of wax).
(a) Showthat θ ≈ 54.7

(assumeh ands areconstant). Hint: Findthecritical point of A(θ) for 0< θ < π/2.
(b) Conﬁrm, bygraphingf (θ) =

3cscθ −cotθ, that thecritical point indeedminimizesthesurfacearea.
71. Findthemaximumof y = x
a
−x
b
on[0, 1] where0< a < b. Inparticular, ﬁndthemaximumof y = x
5
−x
10
on
[0, 1].
solution
• Let f (x) = x
a
− x
b
. Then f

(x) = ax
a−1
− bx
b−1
. Sincea < b, f

(x) = x
a−1
(a − bx
b−a
) = 0 implies
critical pointsx = 0andx = (
a
b
)
1/(b−a)
, whichisintheinterval [0, 1] asa < b implies
a
b
< 1andconsequently
x = (
a
b
)
1/(b−a)
< 1. Also, f (0) = f (1) = 0 and a < b implies x
a
> x
b
on theinterval [0, 1], which gives
f (x) > 0andthusthemaximumvalueof f on[0, 1] is
f
_
_
a
b
_
1/(b−a)
_
=
_
a
b
_
a/(b−a)

_
a
b
_
b/(b−a)
.
• Let f (x) = x
5
−x
10
. Thenbypart (a), themaximumvalueof f on[0, 1] is
f
_
_
1
2
_
1/5
_
=
_
1
2
_

_
1
2
_
2
=
1
2

1
4
=
1
4
.
In Exercises 72–74, plot the function using a graphing utility and ﬁnd its critical points and extreme values on [−5, 5].
y =
1
1+|x −1|
73. y =
1
1+|x −1|
+
1
1+|x −4|
solution Let
f (x) =
1
1+|x −1|
+
1
1+|x −4|
.
Theplot follows:
−5 −4 −3 −2 −1 1 2 3 4 5
0.2
0.4
0.6
0.8
1
1.2
Wecan seeon theplot that thecritical points of f (x) lieat thecusps at x = 1and x = 4and at thelocation of the
horizontal tangent lineat x =
5
2
. Withf (−5) =
17
70
, f (1) = f (4) =
5
4
, f (
5
2
) =
4
5
andf (5) =
7
10
, it followsthat the
maximumvalueof f (x) on[−5, 5] isf (1) = f (4) =
5
4
andtheminimumvalueisf (−5) =
17
70
.
y =
x
|x
2
−1| +|x
2
−4|
75. (a) Useimplicit differentiationtoﬁndthecritical pointsonthecurve27x
2
= (x
2
+y
2
)
3
.
(b) Plot thecurveandthehorizontal tangent linesonthesameset of axes.
solution
(a) Differentiatingbothsidesof theequation27x
2
= (x
2
+y
2
)
3
withrespect tox yields
54x = 3(x
2
+y
2
)
2
_
2x +2y
dy
dx
_
.
Solvingfor dy/dx weobtain
dy
dx
=
27x −3x(x
2
+y
2
)
2
3y(x
2
+y
2
)
2
=
x(9−(x
2
+y
2
)
2
)
y(x
2
+y
2
)
2
.
Thus, thederivativeiszerowhenx
2
+y
2
= 3. Substitutingintotheequationforthecurve, thisyieldsx
2
= 1, orx = ±1.
Therearethereforefour pointsat whichthederivativeiszero:
(−1, −

2), (−1,

2), (1, −

2), (1,

2).
June 9, 2011 LTSV SSM Second Pass
190 C HA P T E R 4 APPLICATIONS OF THE DERIVATIVE
Therearealsocritical pointswherethederivativedoesnot exist. Thisoccurswheny = 0andgivesthefollowingpoints
withvertical tangents:
(0, 0), (±
4

27, 0).
(b) Thecurve27x
2
= (x
2
+y
2
)
3
anditshorizontal tangentsareplottedbelow.
1
−1
y
x
−2 −1 1 2
Sketchthegraphof acontinuousfunctionon(0, 4) withaminimumvaluebut nomaximumvalue.
77. Sketchthegraphof acontinuousfunctionon(0, 4) havingalocal minimumbut noabsoluteminimum.
solution Hereisthegraphof afunctionf on(0, 4) withalocal minimumvalue[betweenx = 2andx = 4] but no
absoluteminimum[sincef (x) →−∞asx →0+].
x
1 2 3
10
−10
y
Sketchthegraphof afunctionon[0, 4] having
(a) Twolocal maximaandonelocal minimum.
(b) Anabsoluteminimumthat occursat anendpoint, andanabsolutemaximumthat occursat acritical point.
79. Sketchthegraphof afunctionf (x) on[0, 4] withadiscontinuity suchthat f (x) hasanabsoluteminimumbut no
absolutemaximum.
solution Hereis thegraphof afunctionf on[0, 4] that (a) has adiscontinuity [at x = 4] and(b) has anabsolute
minimum[at x = 0] but noabsolutemaximum[sincef (x) →∞asx →4−].
y
x
0
1
2
3
4
1 2 3 4
Arainbowisproducedbylightraysthatenteraraindrop(assumedspherical)andexitafterbeingreﬂectedinternally
asinFigure23. Theanglebetweentheincomingandreﬂectedraysisθ = 4r − 2i, wheretheangleof incidencei
andrefractionr arerelatedbySnell’sLawsini = n sinr withn ≈ 1.33(theindexof refractionfor air andwater).
(a) UseSnell’sLawtoshowthat
dr
di
=
cosi
n cosr
.
(b) Showthat themaximumvalueθ
max
of θ occurswheni satisﬁescosi =
_
n
2
−1
3
. Hint: Showthat

di
= 0if
cosi =
n
2
cosr. ThenuseSnell’sLawtoeliminater.
(c) Showthat θ
max
≈ 59.58

.
Further Insights and Challenges
81. Showthat theextremevaluesof f (x) = a sinx +b cosx are±
_
a
2
+b
2
.
solution If f (x) = a sinx +b cosx, thenf

(x) = a cosx −b sinx, sothatf

(x) = 0impliesa cosx −b sinx = 0.
Thisimpliestanx =
a
b
. Then,
sinx =
±a
_
a
2
+b
2
and cosx =
±b
_
a
2
+b
2
.
Therefore
f (x) = a sinx +b cosx = a
±a
_
a
2
+b
2
+b
±b
_
a
2
+b
2
= ±
a
2
+b
2
_
a
2
+b
2
= ±
_
a
2
+b
2
.
Show, byconsideringitsminimum, that f (x) = x
2
−2x +3takesononlypositivevalues. Moregenerally, ﬁnd
theconditionsonr ands under whichthequadraticfunctionf (x) = x
2
+rx +s takesononlypositivevalues. Give
examplesof r ands for whichf takesonbothpositiveandnegativevalues.
83. Show that if thequadratic polynomial f (x) = x
2
+ rx + s takes on both positiveand negativevalues, then its
minimumvalueoccursat themidpoint betweenthetworoots.
solution Let f (x) = x
2
+ rx + s and suppose that f (x) takes on both positive and negative values. This will
guaranteethat f hastworeal roots. Bythequadraticformula, therootsof f are
x =
−r ±
_
r
2
−4s
2
.
June 9, 2011 LTSV SSM Second Pass
S E C T I ON 4.3 The Mean Value Theorem and Monotonicity 191
Observethat themidpoint betweentheserootsis
1
2
_
−r +
_
r
2
−4s
2
+
−r −
_
r
2
−4s
2
_
= −
r
2
.
Next, f

(x) = 2x +r = 0whenx = −
r
2
and, becausethegraphof f (x) isanupwardopeningparabola, it followsthat
f (−
r
2
) isaminimum. Thus, f takesonitsminimumvalueat themidpoint betweenthetworoots.
GeneralizeExercise83: Showthat if thehorizontal liney = c intersectsthegraphof f (x) = x
2
+rx +s at two
points(x
1
, f (x
1
)) and(x
2
, f (x
2
)), thenf (x) takesitsminimumvalueat themidpoint M =
x
1
+x
2
2
(Figure24).
85. A cubic polynomial may have a local min and max, or it may have neither (Figure 25). Find conditions on the
coefﬁcientsa andb of
f (x) =
1
3
x
3
+
1
2
ax
2
+bx +c
that ensurethat f hasneither alocal minnor alocal max. Hint: ApplyExercise82tof

(x).
−4 −2 4 2
(A) (B)
−2 4 2
20
−20
60
30
x x
y y
FIGURE 25 Cubicpolynomials
solution Let f (x) =
1
3
x
3
+
1
2
ax
2
+bx +c. UsingExercise82, wehaveg(x) = f

(x) = x
2
+ax +b > 0for all
x providedb >
1
4
a
2
, inwhichcasef hasnocritical pointsandhencenolocal extrema. (Actuallyb ≥
1
4
a
2
will sufﬁce,
sinceinthiscase[aswe’ll seeinalater section] f hasaninﬂectionpoint but nolocal extrema.)
Findtheminandmaxof
f (x) = x
p
(1−x)
q
on[0, 1],
wherep, q > 0.
87. Provethat if f iscontinuousandf (a) andf (b) arelocal minimawherea < b, thenthereexistsavaluec
betweena andb suchthat f (c) isalocal maximum. (Hint: ApplyTheorem1totheinterval [a, b].) Showthat continuity
isanecessaryhypothesisbysketchingthegraphof afunction(necessarilydiscontinuous) withtwolocal minimabut no
local maximum.
solution
• Let f (x) beacontinuous functionwithf (a) andf (b) local minimaontheinterval [a, b]. By Theorem1, f (x)
musttakeonbothaminimumandamaximumon[a, b]. Sincelocal minimaoccur atf (a) andf (b), themaximum
must occur at someother point intheinterval, call it c, wheref (c) isalocal maximum.
• Thefunctiongraphedhereisdiscontinuousat x = 0.
x
2 4 6 8 −8 −6 −4 −2
4
6
8
y
4.3 The Mean Value Theorem and Monotonicity
Preliminary Questions
1. For whichvalueof m isthefollowingstatement correct?If f (2) = 3andf (4) = 9, andf (x) isdifferentiable, then
f hasatangent lineof slopem.
solution TheMeanValueTheoremguaranteesthat thefunctionhasatangent linewithslopeequal to
f (4) −f (2)
4−2
=
9−3
4−2
= 3.
Hence, m = 3makesthestatement correct.
2. Assumef isdifferentiable. Whichof thefollowingstatementsdoesnot followfromtheMVT?
(a) If f hasasecant lineof slope0, thenf hasatangent lineof slope0.
(b) If f (5) < f (9), thenf

(c) > 0for somec ∈ (5, 9).
June 9, 2011 LTSV SSM Second Pass
192 C HA P T E R 4 APPLICATIONS OF THE DERIVATIVE
(c) If f hasatangent lineof slope0, thenf hasasecant lineof slope0.
(d) If f

(x) > 0for all x, theneverysecant linehaspositiveslope.
solution Conclusion(c)doesnot followfromtheMeanValueTheorem. Asacounterexample, consider thefunction
f (x) = x
3
. Notethat f

(0) = 0, but nosecant linehaszeroslope.
3. Canafunctionthat takesononlynegativevalueshaveapositivederivative? If so, sketchanexample.
solution Yes. Theﬁgurebelowdisplaysafunctionthat takesononlynegativevaluesbut hasapositivederivative.
x
y
4. For f (x) withderivativeasinFigure12:
(a) Isf (c) alocal minimumor maximum?
(b) Isf (x) adecreasingfunction?
c
x
y
FIGURE 12 Graphof derivativef

(x).
solution
(a) Totheleft of x = c, thederivativeispositive, sof isincreasing; totheright of x = c, thederivativeisnegative, sof
isdecreasing. Consequently, f (c) must bealocal maximum.
(b) No. Thederivativeisadecreasingfunction, but asnotedinpart (a), f (x) isincreasingfor x < c anddecreasingfor
x > c.
Exercises
In Exercises 1–8, ﬁnd a point c satisfying the conclusion of the MVT for the given function and interval.
1. y = x
−1
, [2, 8]
solution Let f (x) = x
−1
, a = 2, b = 8. Thenf

(x) = −x
−2
, andbytheMVT, thereexistsac ∈ (2, 8) suchthat

1
c
2
= f

(c) =
f (b) −f (a)
b −a
=
1
8

1
2
8−2
= −
1
16
.
Thusc
2
= 16andc = ±4. Choosec = 4∈ (2, 8).
y =

x, [9, 25]
3. y = cosx −sinx, [0, 2π]
solution Let f (x) = cosx − sinx, a = 0, b = 2π. Thenf

(x) = −sinx − cosx, andby theMVT, thereexistsa
c ∈ (0, 2π) suchthat
−sinc −cosc = f

(c) =
f (b) −f (a)
b −a
=
1−1
2π −0
= 0.
Thus−sinc = cosc. Chooseeither c =

4
or c =

4
∈ (0, 2π).
y =
x
x +2
, [1, 4]
5. y = x
3
, [−4, 5]
solution Let f (x) = x
3
, a = −4, b = 5. Thenf

(x) = 3x
2
, andbytheMVT, thereexistsac ∈ (−4, 5) suchthat
3c
2
= f

(c) =
f (b) −f (a)
b −a
=
189
9
= 21.
Solvingfor c yieldsc
2
= 7, soc = ±

7. Bothof thesevaluesareintheinterval [−4, 5], soeither valuecanbechosen.
y = (x −1)(x −3), [1, 3]
June 9, 2011 LTSV SSM Second Pass
S E C T I ON 4.3 The Mean Value Theorem and Monotonicity 193
7. y = x sinx,
_

π
2
,
π
2
_
solution Let f (x) = x sinx, a = −
π
2
, b =
π
2
. Then f

(x) = sinx + x cosx, and by theMVT, thereexists a
c ∈
_

π
2
,
π
2
_
suchthat
sinc +c cosc = f

(c) =
f (b) −f (a)
b −a
=
π
2

−π
2
(−1)
π
= 0
Solvingfor c givesc = 0astheonlysolution.
y = x −sin(πx), [−1, 1]
9. Let f (x) = x
5
+ x
2
. Thesecant linebetweenx = 0andx = 1has slope2(check this), so by theMVT,
f

(c) = 2for somec ∈ (0, 1). Plot f (x) andthesecant lineonthesameaxes. Thenplot y = 2x +b for different values
of b until thelinebecomestangent tothegraphof f . Zoominonthepoint of tangencytoestimatex-coordinatec of the
point of tangency.
solution Let f (x) = x
5
+x
2
. Theslopeof thesecant linebetweenx = 0andx = 1is
f (1) −f (0)
1−0
=
2−0
1
= 2.
A plot of f (x), thesecant linebetweenx = 0andx = 1, andtheliney = 2x − 0.764isshownbelowat theleft. The
liney = 2x −0.764appearstobetangent tothegraphof y = f (x). Zoominginonthepoint of tangency(seebelowat
theright), it appearsthat thex-coordinateof thepoint of tangencyisapproximately0.62.
y = x
5
+ x
2
y = 2x − .764
x
1
2
4
y
x
y
0.3
0.6
0.5
0.4
0.56 0.6 0.64 0.52
Plotthederivativeof f (x) = 3x
5
−5x
3
. Describeitssignchangesandusethistodeterminethelocal extreme
valuesof f (x). Thengraphf (x) toconﬁrmyour conclusions.
11. Determinetheintervalsonwhichf

(x) ispositiveandnegative, assumingthat Figure13isthegraphof f (x).
x
6 5 4 3 2 1
y
FIGURE 13
solution Thederivativeof f ispositiveontheintervals(0, 1) and(3, 5) wheref isincreasing; it isnegativeonthe
intervals(1, 3) and(5, 6) wheref isdecreasing.
Determinetheintervalsonwhichf (x) isincreasingor decreasing, assumingthatFigure13isthegraphof f

(x).
13. Statewhether f (2) andf (4) arelocal minimaor local maxima, assumingthat Figure13isthegraphof f

(x).
solution
• f

(x) makesatransitionfrompositivetonegativeat x = 2, sof (2) isalocal maximum.
• f

(x) makesatransitionfromnegativetopositiveat x = 4, sof (4) isalocal minimum.
Figure14showsthegraphof thederivativef

(x) of afunctionf (x). Findthecritical pointsof f (x) anddetermine
whether theyarelocal minima, local maxima, or neither.
In Exercises 15–18, sketch the graph of a function f (x) whose derivative f

(x) has the given description.
15. f

(x) > 0for x > 3andf

(x) < 0for x < 3
solution Hereisthegraphof afunctionf for whichf

(x) > 0for x > 3andf

(x) < 0for x < 3.
y
x
0
2
4
6
8
10
1 2 3 4 5
June 9, 2011 LTSV SSM Second Pass
194 C HA P T E R 4 APPLICATIONS OF THE DERIVATIVE
f

(x) > 0for x < 1andf

(x) < 0for x > 1
17. f

(x) isnegativeon(1, 3) andpositiveeverywhereelse.
solution Hereisthegraphof afunctionf for whichf

(x) isnegativeon(1, 3) andpositiveelsewhere.
x
1 2 3 4
2
4
6
8
−2
y
f

(x) makesthesigntransitions+, −, +, −.
In Exercises 19–22, ﬁnd all critical points of f and use the First Derivative Test to determine whether they are local
minima or maxima.
19. f (x) = 4+6x −x
2
solution Let f (x) = 4+6x −x
2
. Thenf

(x) = 6−2x = 0impliesthat x = 3istheonlycritical point of f . Asx
increasesthrough3, f

(x) makesthesigntransition+, −. Therefore, f (3) = 13isalocal maximum.
f (x) = x
3
−12x −4 21. f (x) =
x
2
x +1
solution Let f (x) =
x
2
x +1
. Then
f

(x) =
x(x +2)
(x +1)
2
= 0
impliesthatx = 0andx = −2arecritical points. Notethatx = −1isnotacritical pointbecauseitisnotinthedomainof
f . Asx increasesthrough−2, f

(x) makesthesigntransition+, −sof (−2) = −4isalocal maximum. Asx increases
through0, f

(x) makesthesigntransition−, + sof (0) = 0isalocal minimum.
f (x) = x
3
+x
−3
In Exercises 23–44, ﬁnd the critical points and the intervals on which the function is increasing or decreasing. Use the
First Derivative Test to determine whether the critical point is a local min or max (or neither).
solution Here is a table legend for Exercises 23–44.
SYMBOL MEANING
− The entity is negative on the given interval.
0 The entity is zero at the speciﬁed point.
+ The entity is positive on the given interval.
U The entity is undeﬁned at the speciﬁed point.
f is increasing on the given interval.
f is decreasing on the given interval.
M f has a local maximum at the speciﬁed point.
m f has a local minimum at the speciﬁed point.
¬ There is no local extremum here.
23. y = −x
2
+7x −17
solution Let f (x) = −x
2
+7x −17. Thenf

(x) = 7−2x = 0yieldsthecritical point c =
7
2
.
x
_
−∞,
7
2
_
7/2
_
7
2
, ∞
_
f

+ 0 −
f M
y = 5x
2
+6x −4
25. y = x
3
−12x
2
solution Let f (x) = x
3
−12x
2
. Thenf

(x) = 3x
2
−24x = 3x(x −8) = 0yieldscritical pointsc = 0, 8.
June 9, 2011 LTSV SSM Second Pass
S E C T I ON 4.3 The Mean Value Theorem and Monotonicity 195
x (−∞, 0) 0 (0, 8) 8 (8, ∞)
f

+ 0 − 0 +
f M m
y = x(x −2)
3
27. y = 3x
4
+8x
3
−6x
2
−24x
solution Let f (x) = 3x
4
+8x
3
−6x
2
−24x. Then
f

(x) = 12x
3
+24x
2
−12x −24
= 12x
2
(x +2) −12(x +2) = 12(x +2)(x
2
−1)
= 12(x −1) (x +1) (x +2) = 0
yieldscritical pointsc = −2, −1, 1.
x (−∞, −2) −2 (−2, −1) −1 (−1, 1) 1 (1, ∞)
f

− 0 + 0 − 0 +
f m M m
y = x
2
+(10−x)
2
29. y =
1
3
x
3
+
3
2
x
2
+2x +4
solution Let f (x) =
1
3
x
3
+
3
2
x
2
+2x +4. Thenf

(x) = x
2
+3x +2= (x +1) (x +2) = 0yieldscritical points
c = −2, −1.
x (−∞, −2) −2 (−2, −1) −1 (−1, ∞)
f

+ 0 − 0 +
f M m
y = x
4
+x
3
31. y = x
5
+x
3
+1
solution Let f (x) = x
5
+x
3
+1. Thenf

(x) = 5x
4
+3x
2
= x
2
(5x
2
+3) yieldsasinglecritical point: c = 0.
x (−∞, 0) 0 (0, ∞)
f

+ 0 +
f ¬
y = x
5
+x
3
+x
33. y = x
4
−4x
3/2
(x > 0)
solution Let f (x) = x
4
− 4x
3/2
for x > 0. Thenf

(x) = 4x
3
− 6x
1/2
= 2x
1/2
(2x
5/2
− 3) = 0, whichgivesus
thecritical point c = (
3
2
)
2/5
. (Note: c = 0isnot intheinterval under consideration.)
x
_
0,
_
3
2
_
2/5
_
3
2
2/5
_
_
3
2
_
2/5
, ∞
_
f

− 0 +
f m
y = x
5/2
−x
2
(x > 0)
35. y = x +x
−1
(x > 0)
solution Let f (x) = x +x
−1
for x > 0. Thenf

(x) = 1−x
−2
= 0yieldsthecritical point c = 1. (Note: c = −1
isnot intheinterval under consideration.)
x (0, 1) 1 (1, ∞)
f

− 0 +
f m
June 9, 2011 LTSV SSM Second Pass
196 C HA P T E R 4 APPLICATIONS OF THE DERIVATIVE
y = x
−2
−4x
−1
(x > 0)
37. y =
1
x
2
+1
solution Let f (x) =
_
x
2
+1
_
−1
. Thenf

(x) = −2x
_
x
2
+1
_
−2
= 0yieldscritical point c = 0.
x (−∞, 0) 0 (0, ∞)
f

+ 0 −
f M
y =
2x +1
x
2
+1
39. y =
x
3
x
2
+1
solution Let f (x) =
x
3
x
2
+1
. Then
f

(x) =
(x
2
+1)(3x
2
) −x
3
(2x)
(x
2
+1)
2
=
x
2
(x
2
+3)
(x
2
+1)
2
= 0
yieldsthesinglecritical point c = 0.
x (−∞, 0) 0 (0, ∞)
f

+ 0 +
f ¬
y =
x
3
x
2
−3
41. y = θ +sinθ +cosθ
solution Let f (θ) = θ + sinθ + cosθ. Thenf

(θ) = 1+ cosθ − sinθ = 0yields thecritical points c =
π
2
and
c = π.
θ
_
0,
π
2
_
π
2
_
π
2
, π
_
π (π, 2π)
f

+ 0 − 0 +
f M m
y = sinθ +

3cosθ
43. y = sin
2
θ +sinθ
solution Let f (θ) = sin
2
θ + sinθ. Thenf

(θ) = 2sinθ cosθ + cosθ = cosθ(2sinθ + 1) = 0yieldsthecritical
pointsc =
π
2
,

6
,

2
, and
11π
6
.
θ
_
0,
π
2
_
π
2
_
π
2
,

6
_

6
_

6
,

2
_

2
_

2
,
11π
6
_
11π
6
_
11π
6
, 2π
_
f

+ 0 − 0 + 0 − 0 +
f M m M m
y = θ −2cosθ, [0, 2π]
45. Findtheminimumvalueof f (x) = x
x
for x > 0.
solution Let f (x) = x
x
. By logarithmic differentiation, weknowthat f

(x) = x
x
(1+ lnx). Thus, x =
1
e
is the
onlycritical point. Becausef

(x) < 0for 0< x <
1
e
andf

(x) > 0for x >
1
e
,
f
_
1
e
_
=
_
1
e
_
1/e
≈ 0.692201
istheminimumvalue.
Showthat f (x) = x
2
+bx +c isdecreasingon
_
−∞, −
b
2
_
andincreasingon
_

b
2
, ∞
_
.
47. Showthat f (x) = x
3
−2x
2
+2x isanincreasingfunction. Hint: Findtheminimumvalueof f

(x).
solution Let f (x) = x
3
−2x
2
+2x. For all x, wehave
f

(x) = 3x
2
−4x +2= 3
_
x −
2
3
_
2
+
2
3

2
3
> 0.
Sincef

(x) > 0for all x, thefunctionf iseverywhereincreasing.
Findconditionsona andb that ensurethat f (x) = x
3
+ax +b isincreasingon(−∞, ∞).
June 9, 2011 LTSV SSM Second Pass
S E C T I ON 4.3 The Mean Value Theorem and Monotonicity 197
49. Leth(x) =
x(x
2
−1)
x
2
+1
andsupposethatf

(x) = h(x). Ploth(x) andusetheplottodescribethelocal extrema
andtheincreasing/decreasingbehavior of f (x). Sketchaplausiblegraphfor f (x) itself.
solution Thegraphof h(x) isshownbelowat theleft. Becauseh(x) isnegativefor x < −1andfor 0 < x < 1, it
follows that f (x) is decreasingfor x < −1andfor 0 < x < 1. Similarly, f (x) is increasingfor −1 < x < 0andfor
x > 1becauseh(x) is positiveontheseintervals. Moreover, f (x) has local minimaat x = −1andx = 1andalocal
maximumat x = 0. A plausiblegraphfor f (x) isshownbelowat theright.
x
0.3
0.2
0.1
2 1 −1 −2
−0.2
x
1
0.5
−1
−0.5
h(x) f(x)
1 2 −2 −1
Sammadetwostatementsthat Deborahfounddubious.
(a) “Theaveragevelocityfor mytripwas70mph; at nopoint intimedidmyspeedometer read70mph.”
(b) “A policemanclockedmegoing70mph, but myspeedometer never read65mph.”
Ineachcase, whichtheoremdidDeborahapplytoproveSam’sstatement false: theIntermediateValueTheoremor
theMeanValueTheorem? Explain.
51. Determine where f (x) = (1000− x)
2
+ x
2
is decreasing. Use this to decide which is larger: 800
2
+ 200
2
or
600
2
+400
2
.
solution If f (x) = (1000− x)
2
+ x
2
, then f

(x) = −2(1000− x) + 2x = 4x − 2000. f

(x) < 0 as long as
x < 500. Therefore, 800
2
+200
2
= f (200) > f (400) = 600
2
+400
2
.
Showthat f (x) = 1−|x| satisﬁestheconclusionof theMVT on[a, b] if botha andb arepositiveor negative,
but not if a < 0andb > 0.
53. Whichvaluesof c satisfytheconclusionof theMVT ontheinterval [a, b] if f (x) isalinear function?
solution Letf (x) = px +q, wherep andq areconstants. Thentheslopeof every secantlineandtangentlineof f is
p.Accordingly, consideringtheinterval [a, b], every pointc ∈ (a, b) satisﬁesf

(c) = p =
f (b) −f (a)
b −a
, theconclusion
of theMVT.
Showthat if f (x) isanyquadraticpolynomial, thenthemidpoint c =
a +b
2
satisﬁestheconclusionof theMVT
on[a, b] for anya andb.
55. Supposethat f (0) = 2andf

(x) ≤ 3for x > 0. Apply theMVT to theinterval [0, 4] to provethat f (4) ≤ 14.
Provemoregenerallythat f (x) ≤ 2+3x for all x > 0.
solution TheMVT, appliedtotheinterval [0, 4], guaranteesthat thereexistsac ∈ (0, 4) suchthat
f

(c) =
f (4) −f (0)
4−0
or f (4) −f (0) = 4f

(c).
Becausec > 0, f

(c) ≤ 3, sof (4) −f (0) ≤ 12. Finally, f (4) ≤ f (0) +12= 14.
Moregenerally, let x > 0. TheMVT, appliedtotheinterval [0, x], guaranteesthereexistsac ∈ (0, x) suchthat
f

(c) =
f (x) −f (0)
x −0
or f (x) −f (0) = f

(c)x.
Becausec > 0, f

(c) ≤ 3, sof (x) −f (0) ≤ 3x. Finally, f (x) ≤ f (0) +3x = 3x +2.
Showthat if f (2) = −2andf

(x) ≥ 5for x > 2, thenf (4) ≥ 8.
57. Showthat if f (2) = 5andf

(x) ≥ 10for x > 2, thenf (x) ≥ 10x −15for all x > 2.
solution Let x > 2. TheMVT, appliedtotheinterval [2, x], guaranteesthereexistsac ∈ (2, x) suchthat
f

(c) =
f (x) −f (2)
x −2
or f (x) −f (2) = (x −2)f

(c).
Becausef

(x) ≥ 10, it followsthat f (x) −f (2) ≥ 10(x −2), or f (x) ≥ f (2) +10(x −2) = 10x −15.
Further Insights and Challenges
Showthat acubicfunctionf (x) = x
3
+ax
2
+bx +c isincreasingon(−∞, ∞) if b > a
2
/3.
59. Provethatif f (0) = g(0) andf

(x) ≤ g

(x) forx ≥ 0, thenf (x) ≤ g(x) forall x ≥ 0. Hint: Showthatf (x) −g(x)
isnonincreasing.
solution Let h(x) = f (x) − g(x). By thesumrule, h

(x) = f

(x) − g

(x). Sincef

(x) ≤ g

(x) for all x ≥ 0,
h

(x) ≤ 0for all x ≥ 0. Thisimpliesthat h isnonincreasing. Sinceh(0) = f (0) −g(0) = 0, h(x) ≤ 0for all x ≥ 0(as
h isnonincreasing, it cannot climbabovezero). Hencef (x) −g(x) ≤ 0for all x ≥ 0, andsof (x) ≤ g(x) for x ≥ 0.
UseExercise59toprovethat x ≤ tanx for 0≤ x <
π
2
.
61. UseExercise 59 and theinequality sinx ≤ x for x ≥ 0 (established inTheorem3 of Section 2.6) to provethe
followingassertionsfor all x ≥ 0(eachassertionfollowsfromthepreviousone).
(a) cosx ≥ 1−
1
2
x
2
(b) sinx ≥ x −
1
6
x
3
(c) cosx ≤ 1−
1
2
x
2
+
1
24
x
4
(d) Canyouguessthenext inequalityintheseries?
June 9, 2011 LTSV SSM Second Pass
198 C HA P T E R 4 APPLICATIONS OF THE DERIVATIVE
solution
(a) Weprovethis using Exercise59: Let g(x) = cosx and f (x) = 1−
1
2
x
2
. Then f (0) = g(0) = 1 and g

(x) =
−sinx ≥ −x = f

(x) for x ≥ 0byExercise60. NowapplyExercise59toconcludethat cosx ≥ 1−
1
2
x
2
for x ≥ 0.
(b) Let g(x) = sinx andf (x) = x −
1
6
x
3
. Thenf (0) = g(0) = 0andg

(x) = cosx ≥ 1−
1
2
x
2
= f

(x) for x ≥ 0by
part (a). NowapplyExercise59toconcludethat sinx ≥ x −
1
6
x
3
for x ≥ 0.
(c) Let g(x) = 1−
1
2
x
2
+
1
24
x
4
andf (x) = cosx. Thenf (0) = g(0) = 1andg

(x) = −x +
1
6
x
3
≥ −sinx = f

(x)
for x ≥ 0bypart (b). NowapplyExercise59toconcludethat cosx ≤ 1−
1
2
x
2
+
1
24
x
4
for x ≥ 0.
(d) Thenext inequality intheseries is sinx ≤ x −
1
6
x
3
+
1
120
x
5
, validfor x ≥ 0. Toconstruct (d) from(c), wenote
that thederivativeof sinx is cosx, andlook for apolynomial (whichwecurrently must do by educatedguess) whose
derivativeis 1−
1
2
x
2
+
1
24
x
4
. Weknowthederivativeof x is 1, andthat atermwhosederivativeis −
1
2
x
2
shouldbe
of theformCx
3
.
d
dx
Cx
3
= 3Cx
2
= −
1
2
x
2
, soC = −
1
6
. A termwhosederivativeis
1
24
x
4
shouldbeof theformDx
5
.
Fromthis,
d
dx
Dx
5
= 5Dx
4
=
1
24
x
4
, sothat 5D =
1
24
, or D =
1
120
.
Supposethat f (x) isafunctionsuchthat f (0) = 1andfor all x, f

(x) = f (x) andf (x) > 0(inChapter 7, we
will seethat f (x) istheexponential functione
x
). Provethat for all x ≥ 0(eachassertionfollowsfromtheprevious
one),
(a) f (x) ≥ 1
(b) f (x) ≥ 1+x
(c) f (x) ≥ 1+x +
1
2
x
2
Thenprovebyinductionthat for everywholenumber n andall x ≥ 0,
f (x) ≥ 1+x +
1
2!
x
2
+· · · +
1
n!
x
n
63. Assumethat f

existsandf

(x) = 0for all x. Provethat f (x) = mx +b, wherem = f

(0) andb = f (0).
solution
• Let f

(x) = 0for all x. Thenf

(x) = constant for all x. Sincef

(0) = m, weconcludethat f

(x) = m for all x.
• Let g(x) = f (x) −mx. Theng

(x) = f

(x) −m = m−m = 0whichimpliesthat g(x) = constant for all x and
consequentlyf (x) −mx = constant for all x. Rearrangingthestatement, f (x) = mx +constant. Sincef (0) = b,
weconcludethat f (x) = mx +b for all x.
Deﬁnef (x) = x
3
sin
_
1
x
_
for x = 0andf (0) = 0.
(a) Showthat f

(x) iscontinuousat x = 0andthat x = 0isacritical point of f .
(b) Examinethegraphsof f (x) andf

(x). CantheFirst DerivativeTest beapplied?
(c) Showthat f (0) isneither alocal minnor alocal max.
65. Supposethat f (x) satisﬁesthefollowingequation(anexampleof adifferential equation):
f

(x) = −f (x) 1
(a) Showthat f (x)
2
+f

(x)
2
= f (0)
2
+f

(0)
2
for all x. Hint: Showthat thefunctionontheleft haszeroderivative.
(b) Verifythat sinx andcosx satisfyEq. (1), anddeducethat sin
2
x +cos
2
x = 1.
solution
(a) Let g(x) = f (x)
2
+f

(x)
2
. Then
g

(x) = 2f (x)f

(x) +2f

(x)f

(x) = 2f (x)f

(x) +2f

(x)(−f (x)) = 0,
wherewehaveusedthefactthatf

(x) = −f (x). Becauseg

(0) = 0forall x, g(x) = f (x)
2
+f

(x)
2
mustbeaconstant
function. Inother words, f (x)
2
+f

(x)
2
= C for someconstant C. Todeterminethevalueof C, wecansubstituteany
number for x. Inparticular, for thisproblem, wewant tosubstitutex = 0andﬁndC = f (0)
2
+f

(0)
2
. Hence,
f (x)
2
+f

(x)
2
= f (0)
2
+f

(0)
2
.
(b) Let f (x) = sinx. Thenf

(x) = cosx andf

(x) = −sinx, so f

(x) = −f (x). Next, let f (x) = cosx. Then
f

(x) = −sinx, f

(x) = −cosx, andweagainhavef

(x) = −f (x). Finally, if wetakef (x) = sinx, theresultfrom
part (a) guaranteesthat
sin
2
x +cos
2
x = sin
2
0+cos
2
0= 0+1= 1.
Suppose that functions f and g satisfy Eq. (1) and have the same initial values—that is, f (0) = g(0) and
f

(0) = g

(0). Provethat f (x) = g(x) for all x. Hint: ApplyExercise65(a) tof −g.
67. UseExercise66to prove: f (x) = sinx is theuniquesolutionof Eq. (1) suchthat f (0) = 0andf

(0) = 1; and
g(x) = cosx istheuniquesolutionsuchthatg(0) = 1andg

(0) = 0. Thisresultcanbeusedtodevelopall theproperties
of thetrigonometricfunctions“analytically”—that is, without referencetotriangles.
solution Inpart (b) of Exercise65, it was shownthat f (x) = sinx satisﬁes Eq. (1), andwecandirectly calculate
that f (0) = sin0= 0andf

(0) = cos0= 1. Supposethereisanother function, call it F(x), that satisﬁesEq. (1) with
thesameinitial conditions: F(0) = 0andF

(0) = 1. By Exercise66, it follows that F(x) = sinx for all x. Hence,
f (x) = sinx is theuniquesolutionof Eq. (1) satisfyingf (0) = 0andf

(0) = 1. Theproof that g(x) = cosx is the
uniquesolutionof Eq. (1) satisfyingg(0) = 1andg

(0) = 0iscarriedout inasimilar manner.
4.4 The Shape of a Graph
Preliminary Questions
1. If f isconcaveup, thenf

is(chooseone):
(a) increasing (b) decreasing
solution Thecorrect responseis (a): increasing. If thefunctionis concaveup, thenf

is positive. Sincef

is the
derivativeof f

, it followsthat thederivativeof f

ispositiveandf

must thereforebeincreasing.
June 9, 2011 LTSV SSM Second Pass
S E C T I ON 4.4 The Shape of a Graph 199
2. What conclusioncanyoudrawif f

(c) = 0andf

(c) < 0?
solution If f

(c) = 0andf

(c) < 0, thenf (c) isalocal maximum.
3. Trueor False? If f (c) isalocal min, thenf

(c) must bepositive.
solution False. f

(c) couldbezero.
4. Trueor False? If f

(x) changesfrom+ to− at x = c, thenf hasapoint of inﬂectionat x = c.
solution False. f will haveapoint of inﬂectionat x = c onlyif x = c isinthedomainof f .
Exercises
1. MatchthegraphsinFigure13withthedescription:
(a) f

(x) < 0for all x. (b) f

(x) goesfrom+ to−.
(c) f

(x) > 0for all x. (d) f

(x) goesfrom− to+.
(A) (B) (C) (D)
FIGURE 13
solution
(a) InC, wehavef

(x) < 0for all x.
(b) InA, f

(x) goesfrom+ to−.
(c) InB, wehavef

(x) > 0for all x.
(d) InD, f

(x) goesfrom− to+.
Matcheachstatement withagraphinFigure14that representscompanyproﬁtsasafunctionof time.
(a) Theoutlookisgreat: Thegrowthratekeepsincreasing.
(b) We’relosingmoney, but not asquicklyasbefore.
(c) We’relosingmoney, andit’sgettingworseastimegoeson.
(d) We’redoingwell, but our growthrateislevelingoff.
(e) Businesshadbeencoolingoff, but nowit’spickingup.
(f) Businesshadbeenpickingup, but nowit’scoolingoff.
In Exercises 3–14, determine the intervals on which the function is concave up or down and ﬁnd the points of inﬂection.
3. y = x
2
−4x +3
solution Let f (x) = x
2
− 4x + 3. Thenf

(x) = 2x − 4andf

(x) = 2> 0for all x. Therefore, f isconcaveup
everywhere, andtherearenopointsof inﬂection.
y = t
3
−6t
2
+4
5. y = 10x
3
−x
5
solution Let f (x) = 10x
3
− x
5
. Thenf

(x) = 30x
2
− 5x
4
andf

(x) = 60x − 20x
3
= 20x(3− x
2
). Now, f is
concaveupfor x < −

3andfor 0< x <

3sincef

(x) > 0there. Moreover, f isconcavedownfor −

3< x < 0
andfor x >

3sincef

(x) < 0there. Finally, becausef

(x) changessignat x = 0andat x = ±

3, f (x) hasapoint
of inﬂectionat x = 0andat x = ±

3.
y = 5x
2
+x
4
7. y = θ −2sinθ, [0, 2π]
solution Letf (θ) = θ −2sinθ.Thenf

(θ) = 1−2cosθ andf

(θ) = 2sinθ. Now, f isconcaveupfor0< θ < π
sincef

(θ) > 0there. Moreover, f isconcavedownfor π < θ < 2π sincef

(θ) < 0there. Finally, becausef

(θ)
changessignat θ = π, f (θ) hasapoint of inﬂectionat θ = π.
y = θ +sin
2
θ, [0, π]
9. y = x(x −8

x) (x ≥ 0)
solution Let f (x) = x(x − 8

x) = x
2
− 8x
3/2
. Thenf

(x) = 2x − 12x
1/2
andf

(x) = 2− 6x
−1/2
. Now, f
is concavedownfor 0 < x < 9sincef

(x) < 0there. Moreover, f is concaveupfor x > 9sincef

(x) > 0there.
Finally, becausef

(x) changessignat x = 9, f (x) hasapoint of inﬂectionat x = 9.
y = x
7/2
−35x
2
11. y = (x −2)(1−x
3
)
solution Let f (x) = (x −2)
_
1−x
3
_
= x − x
4
− 2+ 2x
3
. Thenf

(x) = 1− 4x
3
+ 6x
2
andf

(x) = 12x −
12x
2
= 12x(1− x) = 0at x = 0andx = 1. Now, f is concaveupon(0, 1) sincef

(x) > 0there. Moreover, f
isconcavedownon(−∞, 0) ∪ (1, ∞) sincef

(x) < 0there. Finally, becausef

(x) changessignat bothx = 0and
x = 1, f (x) hasapoint of inﬂectionat bothx = 0andx = 1.
y = x
7/5 13. y =
1
x
2
+3
solution Let f (x) =
1
x
2
+3
. Thenf

(x) = −
2x
(x
2
+3)
2
and
f

(x) = −
2(x
2
+3)
2
−8x
2
(x
2
+3)
(x
2
+3)
4
=
6x
2
−6
(x
2
+3)
3
.
Now, f is concaveupfor |x| > 1sincef

(x) > 0there. Moreover, f is concavedownfor |x| < 1sincef

(x) < 0
there. Finally, becausef

(x) changessignat bothx = −1andx = 1, f (x) hasapoint of inﬂectionat bothx = −1and
x = 1.
June 9, 2011 LTSV SSM Second Pass
200 C HA P T E R 4 APPLICATIONS OF THE DERIVATIVE
y =
x
x
2
+9
15. Thegrowthof asunﬂower duringtheﬁrst 100days after sproutingis modeledwell by thelogistic curve
y = h(t ) showninFigure15. Estimatethegrowthrateat thepoint of inﬂectionandexplainitssigniﬁcance. Thenmake
aroughsketchof theﬁrst andsecondderivativesof h(t ).
20 40 60 80 100
50
100
150
200
300
250
t (days)
Height (cm)
FIGURE 15
solution Thepoint of inﬂectioninFigure15appears to occur at t = 40days. Thegraphbelowshows thelogistic
curvewithanapproximatetangent linedrawnat t = 40. Theapproximatetangent linepassesroughlythroughthepoints
(20, 20) and(60, 240). Thegrowthrateat thepoint of inﬂectionisthus
240−20
60−20
=
220
40
= 5.5cm/day.
Becausethelogistic curvechanges fromconcaveup to concavedown at t = 40, thegrowth rateat this point is the
maximumgrowthratefor thesunﬂower plant.
20 40 60 80 100
50
100
150
200
300
250
t (days)
Height (cm)
Sketchesof theﬁrst andsecondderivativeof h(t ) areshownbelowat theleft andat theright, respectively.
20 40 60 80 100
1
2
3
4
6
5
t
100 80 60 40 20
t
h′
0.1
−0.1
h′′
AssumethatFigure16isthegraphof f (x). Wheredothepointsof inﬂectionof f (x) occur, andonwhichinterval
isf (x) concavedown?
17. Repeat Exercise16but assumethat Figure16isthegraphof thederivative f

(x).
solution Points of inﬂectionoccur whenf

(x) changes sign. Consequently, points of inﬂectionoccur whenf

(x)
changesfromincreasingtodecreasingor fromdecreasingtoincreasing. InFigure16, thisoccursat x = b andat x = e;
therefore, f (x) has aninﬂectionpoint at x = b andanother at x = e. Thefunctionf (x) will beconcavedownwhen
f

(x) < 0or whenf

(x) isdecreasing. Thus, f (x) isconcavedownfor b < x < e.
Repeat Exercise16but assumethat Figure16isthegraphof thesecond derivative f

(x).
19. Figure17shows thederivative f

(x) on[0, 1.2]. Locatethepoints of inﬂectionof f (x) andthepoints wherethe
local minimaandmaximaoccur. Determinetheintervalsonwhichf (x) hasthefollowingproperties:
(a) Increasing (b) Decreasing
(c) Concaveup (d) Concavedown
1.2 1 0.17 0.64 0.4
x
y
y = f '(x)
FIGURE 17
June 9, 2011 LTSV SSM Second Pass
S E C T I ON 4.4 The Shape of a Graph 201
solution Recall thatthegraphisthatof f

, not f . Theinﬂectionpointsof f occur wheref

changesfromincreasing
todecreasingor viceversabecauseit isat thesepointsthat thesignof f

changes. Fromthegraphweconcludethat f
has points of inﬂectionat x = 0.17, x = 0.64, andx = 1. Thelocal extremaof f occur wheref

changes sign. This
occursat x = 0.4. Becausethesignof f

changesfrom+ to−, f (0.4) isalocal maximum. Therearenolocal minima.
(a) f isincreasingwhenf

ispositive. Hence, f isincreasingon(0, 0.4).
(b) f isdecreasingwhenf

isnegative. Hence, f isdecreasingon(0.4, 1) ∪ (1, 1.2).
(c) Nowf isconcaveupwheref

isincreasing. Thisoccurson(0, 0.17) ∪ (0.64, 1).
(d) Moreover, f isconcavedownwheref

isdecreasing. Thisoccurson(0.17, 0.64) ∪ (1, 1.2).
Leticiahasbeensellingsolar-poweredlaptopchargersthroughherwebsite, withmonthlysalesasrecordedbelow.
Inareporttoinvestors, shestates, “Salesreachedapointof inﬂectionwhenI startedusingpay-per-clickadvertising.”
Inwhichmonthdidthat occur? Explain.
Month 1 2 3 4 5 6 7 8
Sales 2 30 50 60 90 150 230 340
In Exercises 21–34, ﬁnd the critical points and apply the Second Derivative Test (or state that it fails).
21. f (x) = x
3
−12x
2
+45x
solution Let f (x) = x
3
− 12x
2
+ 45x. Thenf

(x) = 3x
2
− 24x + 45= 3(x − 3)(x − 5), andthecritical points
arex = 3andx = 5. Moreover, f

(x) = 6x −24, sof

(3) = −6< 0andf

(5) = 6> 0. Therefore, bytheSecond
DerivativeTest, f (3) = 54isalocal maximum, andf (5) = 50isalocal minimum.
f (x) = x
4
−8x
2
+1
23. f (x) = 3x
4
−8x
3
+6x
2
solution Let f (x) = 3x
4
− 8x
3
+ 6x
2
. Thenf

(x) = 12x
3
− 24x
2
+ 12x = 12x(x − 1)
2
= 0at x = 0, 1and
f

(x) = 36x
2
− 48x + 12. Thus, f

(0) > 0, whichimpliesf (0) isalocal minimum; however, f

(1) = 0, whichis
inconclusive.
f (x) = x
5
−x
3
25. f (x) =
x
2
−8x
x +1
solution Let f (x) =
x
2
−8x
x +1
. Then
f

(x) =
x
2
+2x −8
(x +1)
2
and f

(x) =
2(x +1)
2
−2(x
2
+2x −8)
(x +1)
3
.
Thus, thecritical points arex = −4 and x = 2. Moreover, f

(−4) < 0 and f

(2) > 0. Therefore, by thesecond
derivativetest, f (−4) = −16isalocal maximumandf (2) = −4isalocal minimum.
f (x) =
1
x
2
−x +2
27. y = 6x
3/2
−4x
1/2
solution Let f (x) = 6x
3/2
− 4x
1/2
. Thenf

(x) = 9x
1/2
− 2x
−1/2
= x
−1/2
(9x − 2), so therearetwo critical
points: x = 0andx =
2
9
. Now,
f

(x) =
9
2
x
−1/2
+x
−3/2
=
1
2
x
−3/2
(9x +2).
Thus, f

_
2
9
_
> 0, whichimpliesf
_
2
9
_
isalocal minimum. f

(x) isundeﬁnedatx = 0, sotheSecondDerivativeTest
cannot beappliedthere.
y = 9x
7/3
−21x
1/2 29. f (x) = x
3
+
48
x
solution Wehavef

(x) = 3x
2
− 48x
−2
, sof

(x) = 0when3x
4
= 48, sox = ±2. Now, f

(x) = 6x + 96x
−3
,
andf

(−2) = −12−
96
8
= −24 < 0whilef

(2) = 12+
96
8
= 24 > 0, so that x = −2is alocal maximumand
x = 2isalocal minimum.
f (x) = x
4
+
128
x
2
31. f (x) = sin
2
x +cosx, [0, π]
solution Let f (x) = sin
2
x +cosx. Thenf

(x) = 2sinx cosx −sinx = sinx(2cosx −1). Ontheinterval [0, π],
f

(x) = 0at x = 0, x =
π
3
andx = π. Now,
f

(x) = 2cos
2
x −2sin
2
x −cosx.
Thus, f

(0) > 0, so f (0) is alocal minimum. Ontheother hand, f

(
π
3
) < 0, so f (
π
3
) is alocal maximum. Finally,
f

(π) > 0, sof (π) isalocal minimum.
y =
1
sinx +4
, [0, 2π]
33. f (x) = 2+tan
2
x,
_

π
2
,
π
2
_
solution Wehavef

(x) = 2tanx sec
2
x andf

(x) = 2sec
4
x + 4tan
2
x sec
2
x. Now, f

(x) = 0inthespeciﬁed
interval whentanx = 0, i.e. whenx = 0. Sincef

(0) = 2, f (0) = 2isalocal minimum.
f (x) = sinx cos
3
x, [0, π]
June 9, 2011 LTSV SSM Second Pass
202 C HA P T E R 4 APPLICATIONS OF THE DERIVATIVE
In Exercises 35–52, ﬁnd the intervals on which f is concave up or down, the points of inﬂection, the critical points, and
the local minima and maxima.
solution Hereisatablelegendfor Exercises35–45.
SYMBOL MEANING
− Theentityisnegativeonthegiveninterval.
0 Theentityiszeroat thespeciﬁedpoint.
+ Theentityispositiveonthegiveninterval.
U Theentityisundeﬁnedat thespeciﬁedpoint.
Thefunction(f , g, etc.) isincreasingonthegiveninterval.
Thefunction(f , g, etc.) isdecreasingonthegiveninterval.
Thefunction(f , g, etc.) isconcaveuponthegiveninterval.
Thefunction(f , g, etc.) isconcavedownonthegiveninterval.
M Thefunction(f , g, etc.) hasalocal maximumat thespeciﬁedpoint.
m Thefunction(f , g, etc.) hasalocal minimumat thespeciﬁedpoint.
I Thefunction(f , g, etc.) hasaninﬂectionpoint here.
¬ Thereisnolocal extremumor inﬂectionpoint here.
35. f (x) = x
3
−2x
2
+x
solution Let f (x) = x
3
−2x
2
+x.
• Thenf

(x) = 3x
2
−4x +1= (x −1)(3x −1) = 0yieldsx = 1andx =
1
3
ascandidatesfor extrema.
• Moreover, f

(x) = 6x −4= 0givesacandidatefor apoint of inﬂectionat x =
2
3
.
x
_
−∞,
1
3
_
1
3
_
1
3
, 1
_
1
_
1, ∞
_
f

+ 0 − 0 +
f M m
x
_
−∞,
2
3
_
2
3
_
2
3
, ∞
_
f

− 0 +
f I
f (x) = x
2
(x −4)
37. f (t ) = t
2
−t
3
solution Let f (t ) = t
2
−t
3
.
• Thenf

(t ) = 2t −3t
2
= t (2−3t ) = 0yieldst = 0andt =
2
3
ascandidatesfor extrema.
• Moreover, f

(t ) = 2−6t = 0givesacandidatefor apoint of inﬂectionat t =
1
3
.
t (−∞, 0) 0
_
0,
2
3
_
2
3
_
2
3
, ∞
_
f

− 0 + 0 −
f m M
t
_
−∞,
1
3
_
1
3
_
1
3
, ∞
_
f

+ 0 −
f I
f (x) = 2x
4
−3x
2
+2
39. f (x) = x
2
−8x
1/2
(x ≥ 0)
solution Let f (x) = x
2
−8x
1/2
. Notethat thedomainof f isx ≥ 0.
• Thenf

(x) = 2x −4x
−1/2
= x
−1/2
_
2x
3/2
−4
_
= 0yieldsx = 0andx = (2)
2/3
ascandidatesfor extrema.
• Moreover, f

(x) = 2+2x
−3/2
> 0for all x ≥ 0, whichmeanstherearenoinﬂectionpoints.
x 0
_
0, (2)
2/3
_
(2)
2/3
_
(2)
2/3
, ∞
_
f

U − 0 +
f M m
f (x) = x
3/2
−4x
−1/2
(x > 0)
June 9, 2011 LTSV SSM Second Pass
S E C T I ON 4.4 The Shape of a Graph 203
41. f (x) =
x
x
2
+27
solution Let f (x) =
x
x
2
+27
.
• Thenf

(x) =
27−x
2
_
x
2
+27
_
2
= 0yieldsx = ±3

3ascandidatesfor extrema.
• Moreover, f

(x) =
−2x
_
x
2
+27
_
2
−(27−x
2
)(2)
_
x
2
+27
_
(2x)
_
x
2
+27
_
4
=
2x
_
x
2
−81
_
_
x
2
+27
_
3
= 0givescandidatesfor
apoint of inﬂectionat x = 0andat x = ±9.
x
_
−∞, −3

3
_
−3

3
_
−3

3, 3

3
_
3

3
_
3

3, ∞
_
f

− 0 + 0 −
f m M
x (−∞, −9) −9 (−9, 0) 0 (0, 9) 9 (9, ∞)
f

− 0 + 0 − 0 +
f I I I
f (x) =
1
x
4
+1
43. f (θ) = θ +sinθ, [0, 2π]
solution Let f (θ) = θ +sinθ on[0, 2π].
• Thenf

(θ) = 1+cosθ = 0yieldsθ = π asacandidatefor anextremum.
• Moreover, f

(θ) = −sinθ = 0givescandidatesfor apoint of inﬂectionat θ = 0, at θ = π, andat θ = 2π.
θ (0, π) π (π, 2π)
f

+ 0 +
f ¬
θ 0 (0, π) π (π, 2π) 2π
f

0 − 0 + 0
f ¬ I ¬
f (x) = cos
2
x, [0, π]
45. f (x) = tanx,
_

π
2
,
π
2
_
solution Let f (x) = tanx on
_

π
2
,
π
2
_
.
• Thenf

(x) = sec
2
x ≥ 1> 0on
_

π
2
,
π
2
_
.
• Moreover, f

(x) = 2secx · secx tanx = 2sec
2
x tanx = 0givesacandidatefor apoint of inﬂectionat x = 0.
x
_

π
2
,
π
2
_
f

+
f
x
_

π
2
, 0
_
0
_
0,
π
2
_
f

− 0 +
f I
f (x) =
x
x
6
+5
x (−∞, −1) −1 (−1, 1) 1 (1, ∞)
f

− 0 + 0 −
f m M
x (−∞, −
6

7) −
6

7 (−
6

7, 0) 0 (0,
6

7)
6

7 (
6

7, ∞)
f

− 0 + 0 − 0 +
f I I I
47. Sketchthegraphof anincreasingfunctionsuchthatf

(x) changesfrom+to−atx = 2andfrom−to+atx = 4.
Dothesamefor adecreasingfunction.
solution Thegraph shown belowat theleft is an increasing function which changes fromconcaveup to concave
downat x = 2andfromconcavedownto concaveupat x = 4. Thegraphshownbelowat theright is adecreasing
functionwhichchangesfromconcaveuptoconcavedownat x = 2andfromconcavedowntoconcaveupat x = 4.
x
2 4
2
1
y
x
2 4
6
2
4
y
June 9, 2011 LTSV SSM Second Pass
204 C HA P T E R 4 APPLICATIONS OF THE DERIVATIVE
In Exercises 48–50, sketch the graph of a function f (x) satisfying all of the given conditions.
f

(x) > 0andf

(x) < 0for all x.
49. (i) f

(x) > 0for all x, and
(ii) f

(x) < 0for x < 0andf

(x) > 0for x > 0.
solution Hereis thegraphof afunctionf (x) satisfying(i) f

(x) > 0for all x and (ii) f

(x) < 0for x < 0and
f

(x) > 0for x > 0.
x
10
5
−10
−5
y
1 2 −2 −1
(i) f

(x) < 0for x < 0andf

(x) > 0for x > 0, and
(ii) f

(x) < 0for |x| > 2, andf

(x) > 0for |x| < 2.
51. Aninfectious ﬂuspreads slowly at thebeginningof anepidemic. Theinfectionprocess accelerates until a
majorityof thesusceptibleindividualsareinfected, at whichpoint theprocessslowsdown.
(a) If R(t ) isthenumber of individualsinfectedat timet , describetheconcavity of thegraphof R near thebeginning
andendof theepidemic.
(b) Describethestatusof theepidemiconthedaythat R(t ) hasapoint of inﬂection.
solution
(a) Near thebeginningof theepidemic, thegraphof R isconcaveup. Near theepidemic’send, R isconcavedown.
(b) “Epidemicsubsiding: number of newcasesdeclining.”
Water ispumpedintoasphereat aconstant rate(Figure18). Let h(t ) bethewater level at timet . Sketchthe
graphof h(t ) (approximately, but withthecorrect concavity). Wheredoesthepoint of inﬂectionoccur?
53. Water is pumped into asphereof radius R at avariableratein such away that thewater level rises at a
constant rate(Figure18). Let V(t ) bethevolumeof water inthetank at timet . SketchthegraphV(t ) (approximately,
but withthecorrect concavity). Wheredoesthepoint of inﬂectionoccur?
solution Becausewater isenteringthesphereinsuchawaythatthewater level risesataconstantrate, weexpectthe
volumetoincreasemoreslowlynear thebottomandtopof thespherewherethesphereisnot as“wide” andtoincrease
morerapidlynear themiddleof thesphere. Thegraphof V(t ) shouldthereforestart concaveupandchangetoconcave
downwhenthesphereishalf full; that is, thepoint of inﬂectionshouldoccur whenthewater level isequal totheradius
of thesphere. A possiblegraphof V(t ) isshownbelow.
t
V
(Continuationof Exercise53) If thespherehasradiusR, thevolumeof water isV = π
_
Rh
2

1
3
h
3
_
whereh is
thewater level. Assumethelevel risesat aconstant rateof 1(that is, h = t ).
(a) Findtheinﬂectionpoint of V(t ). Doesthisagreewithyour conclusioninExercise53?
(b) Plot V(t ) for R = 1.
Further Insights and Challenges
In Exercises 55–57, assume that f (x) is differentiable.
55. Proof of theSecondDerivativeTest Let c beacritical point suchthat f

(c) > 0(thecasef

(c) < 0issimilar).
(a) Showthat f

(c) = lim
h→0
f

(c +h)
h
.
(b) Use(a) toshowthatthereexistsanopeninterval (a, b) containingc suchthatf

(x) < 0if a < x < c andf

(x) > 0
if c < x < b. Concludethat f (c) isalocal minimum.
solution
(a) Becausec isacritical point, either f

(c) = 0or f

(c) doesnotexist; however, f

(c) exists, sof

(c) mustalsoexist.
Therefore, f

(c) = 0. Now, fromthedeﬁnitionof thederivative, wehave
f

(c) = lim
h→0
f

(c +h) −f

(c)
h
= lim
h→0
f

(c +h)
h
.
June 9, 2011 LTSV SSM Second Pass
S E C T I ON 4.4 The Shape of a Graph 205
(b) Wearegiventhat f

(c) > 0. Bypart (a), it followsthat
lim
h→0
f

(c +h)
h
> 0;
inother words, for sufﬁcientlysmall h,
f

(c +h)
h
> 0.
Now, if h issufﬁciently small but negative, thenf

(c + h) must alsobenegative(sothat theratiof

(c + h)/h will be
positive) andc + h < c. Ontheother hand, if h issufﬁciently small but positive, thenf

(c + h) must alsobepositive
andc +h > c. Thus, thereexistsanopeninterval (a, b) containingc suchthat f

(x) < 0for a < x < c andf

(c) > 0
for c < x < b. Finally, becausef

(x) changesfromnegativetopositiveat x = c, f (c) must bealocal minimum.
Provethat if f

(x) existsandf

(x) > 0for all x, thenthegraphof f (x) “sitsabove” itstangent lines.
(a) For anyc, setG(x) = f (x) −f

(c)(x −c) −f (c). ItissufﬁcienttoprovethatG(x) ≥ 0for all c. Explainwhy
withasketch.
(b) Showthat G(c) = G

(c) = 0andG

(x) > 0for all x. Concludethat G

(x) < 0for x < c andG

(x) > 0for
x > c. Thendeduce, usingtheMVT, that G(x) > G(c) for x = c.
57. Assumethat f

(x) existsandlet c beapoint of inﬂectionof f (x).
(a) Usethemethodof Exercise56toprovethat thetangent lineat x = c crosses the graph (Figure19). Hint: Showthat
G(x) changessignat x = c.
(b) Verifythisconclusionfor f (x) =
x
3x
2
+1
bygraphingf (x) andthetangent lineat eachinﬂectionpoint on
thesameset of axes.
FIGURE 19 Tangent linecrossesgraphat point of inﬂection.
solution
(a) LetG(x) = f (x) −f

(c)(x −c) −f (c). Then, asinExercise55, G(c) = G

(c) = 0andG

(x) = f

(x). If f

(x)
changes frompositiveto negativeat x = c, thenso does G

(x) andG

(x) is increasingfor x < c anddecreasingfor
x > c. Thismeansthat G

(x) < 0for x < c andG

(x) < 0for x > c. Thisinturnimpliesthat G(x) isdecreasing, so
G(x) > 0for x < c butG(x) < 0for x > c. Ontheother hand, if f

(x) changesfromnegativetopositiveatx = c, then
sodoesG

(x) andG

(x) isdecreasingfor x < c andincreasingfor x > c. Thus, G

(x) > 0for x < c andG

(x) > 0
for x > c. Thisinturnimpliesthat G(x) isincreasing, soG(x) < 0for x < c andG(x) > 0for x > c. Ineither case,
G(x) changessignat x = c, andthetangent lineat x = c crossesthegraphof thefunction.
(b) Let f (x) =
x
3x
2
+1
. Then
f

(x) =
1−3x
2
(3x
2
+1)
2
and f

(x) =
−18x(1−x
2
)
(3x
2
+1)
3
.
Thereforef (x) hasapoint of inﬂectionat x = 0andat x = ±1. Theﬁgurebelowshowsthegraphof y = f (x) andits
tangentlinesateachof thepointsof inﬂection. Itisclear thateachtangentlinecrossesthegraphof f (x) attheinﬂection
point.
x
y
Let C(x) bethecost of producingx unitsof acertaingood. Assumethat thegraphof C(x) isconcaveup.
(a) ShowthattheaveragecostA(x) = C(x)/x isminimizedattheproductionlevel x
0
suchthataveragecostequals
marginal cost—that is, A(x
0
) = C

(x
0
).
(b) Showthat thelinethrough(0, 0) and(x
0
, C(x
0
)) istangent tothegraphof C(x).
59. Let f (x) beapolynomial of degreen ≥ 2. Showthat f (x) hasat least onepoint of inﬂectionif n isodd. Thengive
anexampletoshowthat f (x) neednot haveapoint of inﬂectionif n iseven.
solution Let f (x) = a
n
x
n
+a
n−1
x
n−1
+· · · +a
1
x +a
0
beapolynomial of degreen. Thenf

(x) = na
n
x
n−1
+
(n −1)a
n−1
x
n−2
+· · · +2a
2
x +a
1
andf

(x) = n(n −1)a
n
x
n−2
+(n −1)(n −2)a
n−1
x
n−3
+· · · +6a
3
x +2a
2
.
If n ≥ 3andisodd, thenn −2isalsooddandf

(x) isapolynomial of odddegree. Thereforef

(x) must takeonboth
positiveandnegativevalues. It followsthat f

(x) hasat least oneroot c suchthat f

(x) changessignat c. Thefunction
f (x) will thenhaveapointof inﬂectionatx = c. Ontheother hand, thefunctionsf (x) = x
2
, x
4
andx
8
arepolynomials
of evendegreethat donot haveanypointsof inﬂection.
Critical andInﬂectionPoints If f

(c) = 0andf (c) isneither alocal minnor alocal max, must x = c bea
point of inﬂection? Thisistruefor “reasonable” functions(includingthefunctionsstudiedinthistext), but it isnot
trueingeneral. Let
f (x)
_
x
2
sin
1
x
for x = 0
June 9, 2011 LTSV SSM Second Pass
206 C HA P T E R 4 APPLICATIONS OF THE DERIVATIVE
4.5 Graph Sketching and Asymptotes
Preliminary Questions
1. Sketchanarcwheref

andf

havethesigncombination++. Dothesamefor −+.
solution Anarc withthesigncombination++ (increasing, concaveup) isshownbelowat theleft. Anarc withthe
signcombination−+ (decreasing, concaveup) isshownbelowat theright.
x
y
x
y
2. If thesigncombinationof f

andf

changesfrom++ to+− at x = c, then(choosethecorrect answer):
(a) f (c) isalocal min (b) f (c) isalocal max
(c) c isapoint of inﬂection
solution Becausethesign of thesecond derivativechanges at x = c, thecorrect responseis (c): c is a point of
inﬂection.
3. Thesecondderivativeof thefunctionf (x) = (x − 4)
−1
isf

(x) = 2(x − 4)
−3
. Althoughf

(x) changessignat
x = 4, f (x) doesnot haveapoint of inﬂectionat x = 4. Whynot?
solution Thefunctionf doesnot haveapoint of inﬂectionat x = 4becausex = 4isnot inthedomainof f.
Exercises
1. Determinethesigncombinationsof f

andf

for eachinterval A–G inFigure15.
C B D E F G A
x
y
y = f (x)
FIGURE 15
solution
• InA, f isdecreasingandconcaveup, sof

< 0andf

> 0.
• InB, f isincreasingandconcaveup, sof

> 0andf

> 0.
• InC, f isincreasingandconcavedown, sof

> 0andf

< 0.
• InD, f isdecreasingandconcavedown, sof

< 0andf

< 0.
• InE, f isdecreasingandconcaveup, sof

< 0andf

> 0.
• InF, f isincreasingandconcaveup, sof

> 0andf

> 0.
• InG, f isincreasingandconcavedown, sof

> 0andf

< 0.
Statethesignchangeat eachtransitionpoint A–G inFigure16. Example: f

(x) goesfrom+ to− at A.
In Exercises 3–6, draw the graph of a function for which f

and f

take on the given sign combinations.
3. ++, +−, −−
solution Thisfunctionchangesfromconcaveuptoconcavedownat x = −1andfromincreasingtodecreasingat
x = 0.
x
y
−1
0 1 −1
June 9, 2011 LTSV SSM Second Pass
S E C T I ON 4.5 Graph Sketching and Asymptotes 207
+−, −−, −+
5. −+, −−, −+
solution Thefunctionisdecreasingeverywhereandchangesfromconcaveuptoconcavedownat x = −1andfrom
concavedowntoconcaveupat x = −
1
2
.
x
y
0.05
−1 0
−+, ++, +−
7. Sketchthegraphof y = x
2
−5x +4.
solution Let f (x) = x
2
− 5x + 4. Thenf

(x) = 2x − 5andf

(x) = 2. Hencef is decreasingfor x < 5/2, is
increasingfor x > 5/2, hasalocal minimumat x = 5/2andisconcaveupeverywhere.
2
5
10
15
4 6
y
x
Sketchthegraphof y = 12−5x −2x
2
.
9. Sketchthegraphof f (x) = x
3
−3x
2
+2. Includethezerosof f (x), whicharex = 1and1±

3(approximately
−0.73, 2.73).
solution Letf (x) = x
3
−3x
2
+2. Thenf

(x) = 3x
2
−6x = 3x(x −2) = 0yieldsx = 0, 2andf

(x) = 6x −6.
Thusf isconcavedownfor x < 1, isconcaveupfor x > 1, hasaninﬂectionpoint at x = 1, isincreasingfor x < 0and
for x > 2, isdecreasingfor 0< x < 2, hasalocal maximumat x = 0, andhasalocal minimumat x = 2.
1
x
1 2 3 −1
y
−1
2
−2
Showthat f (x) = x
3
−3x
2
+6x hasapoint of inﬂectionbut nolocal extremevalues. Sketchthegraph.
11. Extendthesketchof thegraphof f (x) = cosx +
1
2
x inExample4totheinterval [0, 5π].
solution Let f (x) = cosx +
1
2
x. Thenf

(x) = −sinx +
1
2
= 0yields critical points at x =
π
6
,

6
,
13π
6
,
17π
6
,
25π
6
, and
29π
6
. Moreover, f

(x) = −cosx sotherearepointsof inﬂectionat x =
π
2
,

2
,

2
,

2
, and

2
.
2
x
2 4 6 8 10 12 14
y
4
6
0
Sketchthegraphsof y = x
2/3
andy = x
4/3
.
In Exercises 13–34, ﬁnd the transition points, intervals of increase/decrease, concavity, and asymptotic behavior. Then
sketch the graph, with this information indicated.
13. y = x
3
+24x
2
solution Let f (x) = x
3
+24x
2
. Thenf

(x) = 3x
2
+48x = 3x (x +16) andf

(x) = 6x +48. Thisshowsthat f
hascritical pointsat x = 0andx = −16andacandidatefor aninﬂectionpoint at x = −8.
Interval (−∞, −16) (−16, −8) (−8, 0) (0, ∞)
Signsof f

andf

+− −− −+ ++
June 9, 2011 LTSV SSM Second Pass
208 C HA P T E R 4 APPLICATIONS OF THE DERIVATIVE
Thus, thereisalocal maximumat x = −16, alocal minimumat x = 0, andaninﬂectionpoint at x = −8. Moreover,
lim
x→−∞
f (x) = −∞ and lim
x→∞
f (x) = ∞.
Hereisagraphof f withthesetransitionpointshighlightedasinthegraphsinthetextbook.
−20−15 −10 −5 5
1000
2000
3000
y
x
y = x
3
−3x +5
15. y = x
2
−4x
3
solution Let f (x) = x
2
− 4x
3
. Thenf

(x) = 2x − 12x
2
= 2x(1− 6x) andf

(x) = 2− 24x. Critical pointsare
at x = 0andx =
1
6
, andthesolecandidatepoint of inﬂectionisat x =
1
12
.
Interval (−∞, 0) (0,
1
12
) (
1
12
,
1
6
) (
1
6
, ∞)
Signsof f

andf

−+ ++ +− −−
Thus, f (0) isalocal minimum, f (
1
6
) isalocal maximum, andthereisapoint of inﬂectionat x =
1
12
. Moreover,
lim
x→±∞
f (x) = ∞.
Hereisthegraphof f withtransitionpointshighlightedasinthetextbook:
0.04
x
0.2 −0.2
y
−0.04
y =
1
3
x
3
+x
2
+3x
17. y = 4−2x
2
+
1
6
x
4
solution Let f (x) =
1
6
x
4
− 2x
2
+ 4. Thenf

(x) =
2
3
x
3
− 4x =
2
3
x
_
x
2
−6
_
andf

(x) = 2x
2
− 4. Thisshows
that f hascritical pointsat x = 0andx = ±

6andhascandidatesfor pointsof inﬂectionat x = ±

2.
Interval (−∞, −

6) (−

6, −

2) (−

2, 0) (0,

2) (

2,

6) (

6, ∞)
Signsof f

andf

−+ ++ +− −− −+ ++
Thus, f haslocal minimaat x = ±

6, alocal maximumat x = 0, andinﬂectionpointsat x = ±

2. Moreover,
lim
x→±∞
f (x) = ∞.
Hereisagraphof f withtransitionpointshighlighted.
5
x
2 −2
y
10
y = 7x
4
−6x
2
+1
19. y = x
5
+5x
solution Let f (x) = x
5
+ 5x. Thenf

(x) = 5x
4
+ 5= 5(x
4
+ 1) andf

(x) = 20x
3
. f

(x) > 0for all x, sothe
graphhasnocritical pointsandisalwaysincreasing. f

(x) = 0at x = 0. Signanalysesreveal that f

(x) changesfrom
negativetopositiveat x = 0, sothat thegraphof f (x) hasaninﬂectionpoint at (0, 0). Moreover,
lim
x→−∞
f (x) = −∞ and lim
x→∞
f (x) = ∞.
Hereisagraphof f withtransitionpointshighlighted.
June 9, 2011 LTSV SSM Second Pass
S E C T I ON 4.5 Graph Sketching and Asymptotes 209
20
x
2 1 −2 −1
y
−40
40
−20
y = x
5
−15x
3
21. y = x
4
−3x
3
+4x
solution Let f (x) = x
4
− 3x
3
+ 4x. Then f

(x) = 4x
3
− 9x
2
+ 4 = (4x
2
− x − 2)(x − 2) and f

(x) =
12x
2
− 18x = 6x(2x − 3). This shows that f has critical points at x = 2andx =

33
8
andcandidatepoints of
inﬂectionat x = 0andx =
3
2
. Signanalysesreveal that f

(x) changesfromnegativetopositiveat x =
1−

33
8
, from
positiveto negativeat x =
1+

33
8
, andagainfromnegativeto positiveat x = 2. Therefore, f (
1−

33
8
) andf (2) are
local minimaof f (x), andf (
1+

33
8
) isalocal maximum. Further signanalysesreveal thatf

(x) changesfrompositive
tonegativeatx = 0andfromnegativetopositiveatx =
3
2
, sothattherearepointsof inﬂectionbothatx = 0andx =
3
2
.
Moreover,
lim
x→±∞
f (x) = ∞.
Hereisagraphof f (x) withtransitionpointshighlighted.
4
x
2 1 −1
y
6
2
−2
y = x
2
(x −4)
2
23. y = x
7
−14x
6
solution Let f (x) = x
7
− 14x
6
. Then f

(x) = 7x
6
− 84x
5
= 7x
5
(x −12) and f

(x) = 42x
5
− 420x
4
=
42x
4
(x −10). Critical pointsareat x = 0andx = 12, andcandidateinﬂectionpointsareat x = 0andx = 10. Sign
analysesreveal that f

(x) changesfrompositivetonegativeat x = 0andfromnegativetopositiveat x = 12. Therefore
f (0) is alocal maximumand f (12) is alocal minimum. Also, f

(x) changes fromnegativeto positiveat x = 10.
Therefore, thereisapoint of inﬂectionat x = 10. Moreover,
lim
x→−∞
f (x) = −∞ and lim
x→∞
f (x) = ∞.
Hereisagraphof f withtransitionpointshighlighted.
−5 5 10
−5 × 10
6
5 × 10
6
1 × 10
7
y
x
y = x
6
−9x
4
25. y = x −4

x
solution Let f (x) = x − 4

x = x − 4x
1/2
. Thenf

(x) = 1− 2x
−1/2
. This shows that f has critical points at
x = 0(wherethederivativedoesnotexist) andatx = 4(wherethederivativeiszero). Becausef

(x) < 0for 0< x < 4
andf

(x) > 0for x > 4, f (4) is alocal minimum. Nowf

(x) = x
−3/2
> 0for all x > 0, so thegraphis always
concaveup. Moreover,
lim
x→∞
f (x) = ∞.
Hereisagraphof f withtransitionpointshighlighted.
−2
2
4
5 10 15 20
y
x
June 9, 2011 LTSV SSM Second Pass
210 C HA P T E R 4 APPLICATIONS OF THE DERIVATIVE
y =

x +

16−x
27. y = x(8−x)
1/3
solution Let f (x) = x (8−x)
1/3
. Then
f

(x) = x ·
1
3
(8−x)
−2/3
(−1) +(8−x)
1/3
· 1=
24−4x
3(8−x)
2/3
andsimilarly
f

(x) =
4x −48
9(8−x)
5/3
.
Critical pointsareat x = 8andx = 6, andcandidateinﬂectionpointsarex = 8andx = 12. Signanalysesreveal that
f

(x) changesfrompositivetonegativeat x = 6andf

(x) remainsnegativeoneither sideof x = 8. Moreover, f

(x)
changesfromnegativetopositiveat x = 8andfrompositivetonegativeat x = 12. Therefore, f hasalocal maximum
at x = 6andinﬂectionpointsat x = 8andx = 12. Moreover,
lim
x→±∞
f (x) = −∞.
Hereisagraphof f withthetransitionpointshighlighted.
−30
−5 5 10 15
−20
−10
y
x
y = (x
2
−4x)
1/3
29. y = (2x −x
2
)
1/3
solution Wehave
f

(x) =
1
3
(2x −x
2
)
−2/3
(2−2x) =
2−2x
3(2x −x
2
)
2/3
f

(x) =
2(x
2
−2x +4)
9x(x −2)(2x −x
2
)
2/3
Theonlycritical pointisatx = 1, andsincethenumerator of f

(x) isalwayspositive, therearenoinﬂectionpoints. Sign
analysisshowsthat f

(x) changesfrompositivetonegativeat x = 1, sothat f hasalocal maximumat x = 1. Finally,
lim
x→±∞
f (x) = ∞
sothat thegraphhasnohorizontal asymptotes. Hereisagraphof f withthetransitionpointshighlighted:
−1
−10 −5 5 10
x
y
1
−2
−3
−4
y = (x
3
−3x)
1/3
31. y = x −x
−1
solution Wehavef

(x) = 1+ x
−2
andf

(x) = −2x
−3
. Sincef (x) never vanishes, therearenocritical points.
Sincef

(x) = 0onlyat x = 0, therearenoinﬂectionpoints. Finally,
lim
x→±∞
f (x) = ±∞, lim
x→0+
f (x) = −∞, lim
x→0−
f (x) = ∞
sothat thegraphhasavertical asymptoteat x = 0but nohorizontal asymptotes. Hereisagraphof f :
June 9, 2011 LTSV SSM Second Pass
S E C T I ON 4.5 Graph Sketching and Asymptotes 211
−3 −1 1 2 3
x
y
−2
4
6
8
−4
−6
−8
y = x
2
−x
−2 33. y = x
3

48
x
2
solution Wehavef

(x) = 3x
2
+ 96x
−3
andf

(x) = 6x − 288x
−4
. Thecritical points occur for 0 = f

(x) =
3x
2
+ 96x
−3
, sowhere3x
5
= −96. Thustheonly critical point isx = −2. f

(x) vanisheswhen6x
5
= 288, sothat
x
5
= 48andx =
5

48. Signanalysisshowsthat f

(x) changesfrompositivetonegativeat x = −2, sothat x = −2isa
local maximum, but doesnot changesign, remainingpositive, at x = 2. Thusx = 2isnot alocal extremum. Moreover,
f

(x) changesfromnegativetopositiveat x =
5

48, sothat thisisapoint of inﬂection. Finally,
lim
x→±∞
f (x) = ±∞, lim
x→0
f (x) = −∞
sothat thegraphhasavertical asymptoteat x = 0but nohorizontal asymptotes. Hereisagraphof f withthetransition
pointshighlighted:
−4 −3 1 2 3 4
y
x
−100
−80
−60
−40
−20
20
40
60
y = x
2
−x +x
−1
35. Sketchthegraphof f (x) = 18(x −3)(x −1)
2/3
usingtheformulas
f

(x) =
30
_
x −
9
5
_
(x −1)
1/3
, f

(x) =
20
_
x −
3
5
_
(x −1)
4/3
solution
f

(x) =
30(x −
9
5
)
(x −1)
1/3
yieldscritical pointsat x =
9
5
, x = 1.
f

(x) =
20(x −
3
5
)
(x −1)
4/3
yieldspotential inﬂectionpointsat x =
3
5
, x = 1.
Interval signsof f

andf

(−∞,
3
5
) +−
(
3
5
, 1) ++
(1,
9
5
) −+
(
9
5
, ∞) ++
Thegraph has an inﬂection point at x =
3
5
, alocal maximumat x = 1 (at which thegraph has acusp), and alocal
minimumat x =
9
5
. Thesketchlookssomethinglikethis.
40
x
3 2 1 −2 −1
y
−20
20
−40
−60
−80
June 9, 2011 LTSV SSM Second Pass
212 C HA P T E R 4 APPLICATIONS OF THE DERIVATIVE
Sketchthegraphof f (x) =
x
x
2
+1
usingtheformulas
f

(x) =
1−x
2
(1+x
2
)
2
, f

(x) =
2x(x
2
−3)
(x
2
+1)
3
In Exercises 37–40, sketch the graph of the function, indicating all transition points. If necessary, use a graphing
utility or computer algebra system to locate the transition points numerically.
37. y = x
3

4
x
2
+1
solution Wehave
f

(x) = 3x
2
+
8x
(x
2
+1)
2
, f

(x) = 6x −
8(3x
2
−1)
(x
2
+1)
3
Thecritical points aretheroots of f

(x), which arex = 0 and x = −0.8678. Candidates for inﬂection points occur
when f

(x) vanishes; theonly root of f

(x) is −0.41119. Sign analysis reveals that f

(x) changes frompositiveto
negativeat x = −0.8678andfromnegativeto positiveat x = 0. Also, f

(x) changes signfromnegativeto positive
at x = −0.41119. Thus f has alocal maximumat x = −0.8678, apoint of inﬂectionat x = −0.41119, andalocal
minimumat x = 0. Moreover, lim
x→±∞
= ±∞, sotherearenohorizontal asymptotes. Sincef (x) isdeﬁnedeverywhere,
therearenovertical asymptotes. Hereisagraphof f withthetransitionpointshighlighted:
−2
−2
2
−6
−1 1 2
x
y
y = 12
_
x
2
+2x +4−x
2
39. y = x
4
−4x
2
+x +1
solution Let f (x) = x
4
− 4x
2
+ x + 1. Thenf

(x) = 4x
3
− 8x + 1andf

(x) = 12x
2
− 8. Thecritical points
arex = −1.473, x = 0.126andx = 1.347, whilethecandidates for points of inﬂectionarex = ±
_
2
3
. Signanalysis
reveals that f

(x) changes fromnegativeto positiveat x = −1.473, frompositiveto negativeat x = 0.126andfrom
negativetopositiveat x = 1.347. For thesecondderivative, f

(x) changesfrompositivetonegativeat x = −
_
2
3
and
fromnegativetopositiveat x =
_
2
3
. Therefore, f haslocal minimaat x = −1.473andx = 1.347, alocal maximumat
x = 0.126andpointsof inﬂectionat x = ±
_
2
3
. Moreover,
lim
x→±∞
f (x) = ∞.
Hereisagraphof f withthetransitionpointshighlighted.
−2 −1 1 2
−5
5
10
15
20
y
x
y = 2

x −sinx, 0≤ x ≤ 2π
In Exercises 41–46, sketch the graph over the given interval, with all transition points indicated.
41. y = x +sinx, [0, 2π]
solution Let f (x) = x +sinx. Settingf

(x) = 1+cosx = 0yieldscosx = −1, sothat x = π isthelonecritical
point ontheinterval [0, 2π]. Settingf

(x) = −sinx = 0yields potential points of inﬂectionat x = 0, π, 2π onthe
interval [0, 2π].
Interval signsof f

andf

(0, π) +−
(π, 2π) ++
Thegraphhasaninﬂectionpoint at x = π, andnolocal maximaor minima. Hereisasketchof thegraphof f (x):
June 9, 2011 LTSV SSM Second Pass
S E C T I ON 4.5 Graph Sketching and Asymptotes 213
y
x
1
2
3
4
5
6
0 6 5 4 3 2 1
y = sinx +cosx, [0, 2π]
43. y = 2sinx −cos
2
x, [0, 2π]
solution Let f (x) = 2sinx − cos
2
x. Thenf

(x) = 2cosx − 2cosx (−sinx) = sin2x + 2cosx andf

(x) =
2cos2x − 2sinx. Settingf

(x) = 0yieldssin2x = −2cosx, sothat 2sinx cosx = −2cosx. Thisimpliescosx = 0
or sinx = −1, sothat x =
π
2
or

2
. Settingf

(x) = 0yields2cos2x = 2sinx, sothat 2sin(
π
2
− 2x) = 2sinx, or
π
2
−2x = x ±2nπ. Thisyields3x =
π
2
+2nπ, or x =
π
6
,

6
,

6
=

2
.
Interval signsof f

andf

_
0,
π
6
_
++
_
π
6
,
π
2
_
+−
_
π
2
,

6
_
−−
_

6
,

2
_
−+
_

2
, 2π
_
++
Thegraphhas alocal maximumat x =
π
2
, alocal minimumat x =

2
, andinﬂectionpoints at x =
π
6
andx =

6
.
Hereisagraphof f without transitionpointshighlighted.
x
6 5 4
3
2 1
y
1
2
−2
−1
y = sinx +
1
2
x, [0, 2π]
45. y = sinx +

3cosx, [0, π]
solution Let f (x) = sinx +

3cosx. Settingf

(x) = cosx −

3sinx = 0yields tanx =
1

3
. Intheinterval
[0, π], thesolutionisx =
π
6
. Settingf

(x) = −sinx −

3cosx = 0yieldstanx = −

3. Intheinterval [0, π], the
lonesolutionisx =

3
.
Interval signsof f

andf

(0, π/6) +−
(π/6, 2π/3) −−
(2π/3, π) −+
Thegraph has alocal maximumat x =
π
6
and apoint of inﬂection at x =

3
. A plot without thetransition points
highlightedisgivenbelow:
x
3 2 1
y
1
2
−2
−1
June 9, 2011 LTSV SSM Second Pass
214 C HA P T E R 4 APPLICATIONS OF THE DERIVATIVE
y = sinx −
1
2
sin2x, [0, π]
47. Areall signtransitionspossible? Explainwithasketchwhythetransitions++ →−+ and−− →+− do
not occur if thefunctionisdifferentiable. (SeeExercise76for aproof.)
solution Inbothcases, thereisapoint wheref isnot differentiableat thetransitionfromincreasingtodecreasingor
decreasingtoincreasing.
y
x
y
x
Supposethat f istwicedifferentiablesatisfying(i) f (0) = 1, (ii) f

(x) > 0for all x = 0, and(iii) f

(x) < 0
for x < 0andf

(x) > 0for x > 0. Let g(x) = f (x
2
).
(a) Sketchapossiblegraphof f (x).
(b) Provethatg(x) hasnopointsof inﬂectionandauniquelocal extremevalueatx = 0. Sketchapossiblegraphof
g(x).
49. Whichof thegraphsinFigure17cannot bethegraphof apolynomial? Explain.
(A) (B) (C)
x
x
x
y y y
FIGURE 17
solution Polynomials areeverywheredifferentiable. Accordingly, graph (B) cannot bethegraph of apolynomial,
sincethefunctionin(B) hasacusp(sharpcorner), signifyingnondifferentiabilityat that point.
WhichcurveinFigure18isthegraphof f (x) =
2x
4
−1
1+x
4
? Explainonthebasisof horizontal asymptotes.
51. MatchthegraphsinFigure19withthetwofunctionsy =
3x
x
2
−1
andy =
3x
2
x
2
−1
. Explain.
(A) (B)
−1 1 −1 1
x x
y y
FIGURE 19
solution Since lim
x→±∞
3x
2
x
2
−1
=
3
1
· lim
x→±∞
1= 3, thegraphof y =
3x
2
x
2
−1
hasahorizontal asymptoteof y = 3;
hence, theright curveisthegraphof f (x) =
3x
2
x
2
−1
. Since
lim
x→±∞
3x
x
2
−1
=
3
1
· lim
x→±∞
x
−1
= 0,
thegraphof y =
3x
x
2
−1
hasahorizontal asymptoteof y = 0; hence, theleft curveisthegraphof f (x) =
3x
x
2
−1
.
Matchthefunctionswiththeir graphsinFigure20.
(a) y =
1
x
2
−1
(b) y =
x
2
x
2
+1
(c) y =
1
x
2
+1
(d) y =
x
x
2
−1
In Exercises 53–70, sketch the graph of the function. Indicate the transition points and asymptotes.
53. y =
1
3x −1
solution Let f (x) =
1
3x −1
. Thenf

(x) =
−3
(3x −1)
2
, sothat f isdecreasingfor all x =
1
3
. Moreover, f

(x) =
18
(3x −1)
3
, so that f is concaveup for x >
1
3
and concavedown for x <
1
3
. Because lim
x→±∞
1
3x −1
= 0, f has a
horizontal asymptoteat y = 0. Finally, f hasavertical asymptoteat x =
1
3
with
lim
x→
1
3

1
3x −1
= −∞ and lim
x→
1
3
+
1
3x −1
= ∞.
June 9, 2011 LTSV SSM Second Pass
S E C T I ON 4.5 Graph Sketching and Asymptotes 215
−5
−2 2
5
y
x
y =
x −2
x −3
55. y =
x +3
x −2
solution Let f (x) =
x +3
x −2
. Thenf

(x) =
−5
(x −2)
2
, so that f is decreasingfor all x = 2. Moreover, f

(x) =
10
(x −2)
3
, sothatf isconcaveupfor x > 2andconcavedownfor x < 2. Because lim
x→±∞
x +3
x −2
= 1, f hasahorizontal
asymptoteat y = 1. Finally, f hasavertical asymptoteat x = 2with
lim
x→2−
x +3
x −2
= −∞ and lim
x→2+
x +3
x −2
= ∞.
x
10 5 −10 −5
y
−10
−5
10
5
y = x +
1
x
57. y =
1
x
+
1
x −1
solution Letf (x) =
1
x
+
1
x −1
. Thenf

(x) = −
2x
2
−2x +1
x
2
(x −1)
2
, sothatf isdecreasingfor all x = 0, 1. Moreover,
f

(x) =
2
_
2x
3
−3x
2
+3x −1
_
x
3
(x −1)
3
, so that f is concaveupfor 0 < x <
1
2
andx > 1andconcavedownfor x < 0
and
1
2
< x < 1. Because lim
x→±∞
_
1
x
+
1
x −1
_
= 0, f has ahorizontal asymptoteat y = 0. Finally, f has vertical
asymptotesat x = 0andx = 1with
lim
x→0−
_
1
x
+
1
x −1
_
= −∞ and lim
x→0+
_
1
x
+
1
x −1
_
= ∞
and
lim
x→1−
_
1
x
+
1
x −1
_
= −∞ and lim
x→1+
_
1
x
+
1
x −1
_
= ∞.
x
1 2 −1
y
5
−5
y =
1
x

1
x −1
59. y =
1
x(x −2)
solution Let f (x) =
1
x(x −2)
. Thenf

(x) =
2(1−x)
x
2
(x −2)
2
, so that f is increasingfor x < 0and0 < x < 1and
decreasingfor 1< x < 2andx > 2. Moreover, f

(x) =
2(3x
2
−6x +4)
x
3
(x −2)
3
, sothatf isconcaveupfor x < 0andx > 2
andconcavedownfor 0< x < 2. Because lim
x→±∞
_
1
x(x −2)
_
= 0, f hasahorizontal asymptoteat y = 0. Finally, f
hasvertical asymptotesat x = 0andx = 2with
lim
x→0−
_
1
x(x −2)
_
= +∞ and lim
x→0+
_
1
x(x −2)
_
= −∞
June 9, 2011 LTSV SSM Second Pass
216 C HA P T E R 4 APPLICATIONS OF THE DERIVATIVE
and
lim
x→2−
_
1
x(x −2)
_
= −∞ and lim
x→2+
_
1
x(x −2)
_
= ∞.
−5
5
−2 4
y
x
y =
x
x
2
−9
61. y =
1
x
2
−6x +8
solution Let f (x) =
1
x
2
−6x +8
=
1
(x −2) (x −4)
. Then f

(x) =
6−2x
_
x
2
−6x +8
_
2
, so that f is increasing for
x < 2andfor 2 < x < 3, is decreasingfor 3 < x < 4andfor x > 4, andhas alocal maximumat x = 3. Moreover,
f

(x) =
2
_
3x
2
−18x +28
_
_
x
2
−6x +8
_
3
, sothat f is concaveupfor x < 2andfor x > 4andis concavedownfor 2 < x < 4.
Because lim
x→±∞
1
x
2
−6x +8
= 0, f hasahorizontal asymptoteat y = 0. Finally, f hasvertical asymptotesat x = 2
andx = 4, with
lim
x→2−
_
1
x
2
−6x +8
_
= ∞ and lim
x→2+
_
1
x
2
−6x +8
_
= −∞
and
lim
x→4−
_
1
x
2
−6x +8
_
= −∞ and lim
x→4+
_
1
x
2
−6x +8
_
= ∞.
x
6 5 4 2
3
1
y
5
−5
y =
x
3
+1
x
63. y = 1−
3
x
+
4
x
3
solution Let f (x) = 1−
3
x
+
4
x
3
. Then
f

(x) =
3
x
2

12
x
4
=
3(x −2)(x +2)
x
4
,
sothat f isincreasingfor |x| > 2anddecreasingfor −2< x < 0andfor 0< x < 2. Moreover,
f

(x) = −
6
x
3
+
48
x
5
=
6(8−x
2
)
x
5
,
so that f is concave down for −2

2 < x < 0 and for x > 2

2, while f is concave up for x < −2

2 and for
0< x < 2

2. Because
lim
x→±∞
_
1−
3
x
+
4
x
3
_
= 1,
f hasahorizontal asymptoteat y = 1. Finally, f hasavertical asymptoteat x = 0with
lim
x→0−
_
1−
3
x
+
4
x
3
_
= −∞ and lim
x→0+
_
1−
3
x
+
4
x
3
_
= ∞.
June 9, 2011 LTSV SSM Second Pass
S E C T I ON 4.5 Graph Sketching and Asymptotes 217
2
x
2 4 6 −6 −4 −2
y
4
6
−6
−4
−2
y =
1
x
2
+
1
(x −2)
2
65. y =
1
x
2

1
(x −2)
2
solution Letf (x) =
1
x
2

1
(x −2)
2
. Thenf

(x) = −2x
−3
+2(x −2)
−3
, sothatf isincreasingfor x < 0andfor
x > 2andisdecreasingfor 0< x < 2. Moreover,
f

(x) = 6x
−4
−6(x −2)
−4
= −
48(x −1)(x
2
−2x +2)
x
4
(x −2)
4
,
sothat f is concaveupfor x < 0andfor 0 < x < 1, is concavedownfor 1 < x < 2andfor x > 2, andhas apoint
of inﬂectionat x = 1. Because lim
x→±∞
_
1
x
2

1
(x −2)
2
_
= 0, f has ahorizontal asymptoteat y = 0. Finally, f has
vertical asymptotesat x = 0andx = 2with
lim
x→0−
_
1
x
2

1
(x −2)
2
_
= ∞ and lim
x→0+
_
1
x
2

1
(x −2)
2
_
= ∞
and
lim
x→2−
_
1
x
2

1
(x −2)
2
_
= −∞ and lim
x→2+
_
1
x
2

1
(x −2)
2
_
= −∞.
x
2
4
−2
−4
y
1 2 3 4
−2 −1
y =
4
x
2
−9
67. y =
1
(x
2
+1)
2
solution Letf (x) =
1
(x
2
+1)
2
. Thenf

(x) =
−4x
(x
2
+1)
3
, sothatf isincreasingfor x < 0, isdecreasingfor x > 0
andhasalocal maximumat x = 0. Moreover,
f

(x) =
−4(x
2
+1)
3
+4x · 3(x
2
+1)
2
· 2x
(x
2
+1)
6
=
20x
2
−4
(x
2
+1)
4
,
sothat f isconcaveupfor |x| > 1/

5, isconcavedownfor |x| < 1/

5, andhaspointsof inﬂectionat x = ±1/

5.
Because lim
x→±∞
1
(x
2
+1)
2
= 0, f hasahorizontal asymptoteat y = 0. Finally, f hasnovertical asymptotes.
x
1
0.8
4 2 −2 −4
y
June 9, 2011 LTSV SSM Second Pass
218 C HA P T E R 4 APPLICATIONS OF THE DERIVATIVE
y =
x
2
(x
2
−1)(x
2
+1)
69. y =
1
_
x
2
+1
solution Let f (x) =
1

x
2
+1
. Then
f

(x) = −
x
_
(x
2
+1)
3
= −x(x
2
+1)
−3/2
,
sothat f isincreasingfor x < 0anddecreasingfor x > 0. Moreover,
f

(x) = −
3
2
x(x
2
+1)
−5/2
(−2x) −(x
2
+1)
−3/2
= (2x
2
−1)(x
2
+1)
−5/2
,
sothat f isconcavedownfor |x| <

2
2
andconcaveupfor |x| >

2
2
. Because
lim
x→±∞
1
_
x
2
+1
= 0,
f hasahorizontal asymptoteat y = 0. Finally, f hasnovertical asymptotes.
x
1
0.8
0.2
10 5 −5 −10
y
y =
x
_
x
2
+1
Further Insights and Challenges
In Exercises 71–75, we explore functions whose graphs approach a nonhorizontal line as x →∞. A line y = ax +b is
called a slant asymptoteif
lim
x→∞
(f (x) −(ax +b)) = 0
or
lim
x→−∞
(f (x) −(ax +b)) = 0
71. Let f (x) =
x
2
x −1
(Figure21). Verifythefollowing:
(a) f (0) isalocal maxandf (2) alocal min.
(b) f isconcavedownon(−∞, 1) andconcaveupon(1, ∞).
(c) lim
x→1−
f (x) = −∞and lim
x→1+
f (x) = ∞.
(d) y = x +1isaslant asymptoteof f (x) asx →±∞.
(e) Theslant asymptoteliesabovethegraphof f (x) for x < 1andbelowthegraphfor x > 1.
y = x + 1
10 −10
10
−10
x
y
f (x) =
x
2
x − 1
FIGURE 21
solution Let f (x) =
x
2
x −1
. Thenf

(x) =
x(x −2)
(x −1)
2
andf

(x) =
2
(x −1)
3
.
(a) Signanalysisof f

(x) revealsthat f

(x) < 0on(−∞, 1) andf

(x) > 0on(1, ∞).
(b) Critical pointsof f

(x) occur at x = 0andx = 2. x = 1isnot acritical point becauseit isnot inthedomainof f .
Signanalysesreveal that x = 2isalocal minimumof f andx = 0isalocal maximum.
June 9, 2011 LTSV SSM Second Pass
S E C T I ON 4.5 Graph Sketching and Asymptotes 219
(c)
lim
x→1−
f (x) = −1 lim
x→1−
1
1−x
= −∞ and lim
x→1+
f (x) = 1 lim
x→1+
1
x −1
= ∞.
(d) Note that using polynomial division, f (x) =
x
2
x −1
= x + 1 +
1
x −1
. Then
lim
x→±∞
(f (x) −(x +1)) = lim
x→±∞
x +1+
1
x −1
−(x +1) = lim
x→±∞
1
x −1
= 0.
(e) For x > 1, f (x) −(x +1) =
1
x −1
> 0, sof (x) approachesx +1fromabove. Similarly, for x < 1, f (x) −(x +
1) =
1
x −1
< 0, sof (x) approachesx +1frombelow.
If f (x) = P(x)/Q(x), whereP andQ arepolynomials of degrees m + 1andm, thenby longdivision, we
canwrite
f (x) = (ax +b) +P
1
(x)/Q(x)
whereP
1
isapolynomial of degree< m. Showthat y = ax +b istheslant asymptoteof f (x). Usethisprocedure
toﬁndtheslant asymptotesof thefollowingfunctions:
(a) y =
x
2
x +2
(b) y =
x
3
+x
x
2
+x +1
73. Sketchthegraphof
f (x) =
x
2
x +1
.
Proceedasinthepreviousexercisetoﬁndtheslant asymptote.
solution Let f (x) =
x
2
x +1
. Thenf

(x) =
x(x +2)
(x +1)
2
andf

(x) =
2
(x +1)
3
. Thus, f isincreasingfor x < −2and
for x > 0, isdecreasingfor −2< x < −1andfor −1< x < 0, hasalocal minimumat x = 0, hasalocal maximumat
x = −2, isconcavedownon(−∞, −1) andconcaveupon(−1, ∞). Limitanalysesgiveavertical asymptoteatx = −1,
with
lim
x→−1−
x
2
x +1
= −∞ and lim
x→−1+
x
2
x +1
= ∞.
Bypolynomial division, f (x) = x −1+
1
x +1
and
lim
x→±∞
_
x −1+
1
x +1
−(x −1)
_
= 0,
whichimpliesthat theslant asymptoteisy = x −1. Noticethat f approachestheslant asymptoteasinexercise71.
x
4
2
4 2 −2 −4
−4
−6
−2
y
Show that y = 3x is aslant asymptotefor f (x) = 3x +x
−2
. Determinewhether f (x) approaches theslant
asymptotefromaboveor belowandmakeasketchof thegraph.
75. Sketchthegraphof f (x) =
1−x
2
2−x
.
solution Let f (x) =
1−x
2
2−x
. Usingpolynomial division, f (x) = x +2+
3
x −2
. Then
lim
x→±∞
(f (x) −(x +2)) = lim
x→±∞
_
(x +2) +
3
x −2
−(x +2)
_
= lim
x→±∞
3
x −2
=
3
1
· lim
x→±∞
x
−1
= 0
which implies that y = x + 2is theslant asymptoteof f (x). Sincef (x) − (x + 2) =
3
x −2
> 0for x > 2, f (x)
approaches theslant asymptotefromabovefor x > 2; similarly,
3
x −2
< 0for x < 2so f (x) approaches theslant
asymptotefrombelowfor x < 2. Moreover, f

(x) =
x
2
−4x +1
(2−x)
2
andf

(x) =
−6
(2−x)
3
. Signanalysesreveal alocal
minimumat x = 2+

3, alocal maximumat x = 2−

3andthat f isconcavedownon(−∞, 2) andconcaveupon
(2, ∞). Limit analysesgiveavertical asymptoteat x = 2.
June 9, 2011 LTSV SSM Second Pass
220 C HA P T E R 4 APPLICATIONS OF THE DERIVATIVE
x
10
5
10 5 −5 −10
−10
−5
y
Assumethat f

(x) andf

(x) exist for all x andlet c beacritical point of f (x). Showthat f (x) cannot makea
transitionfrom++ to−+ at x = c. Hint: ApplytheMVT tof

(x).
77. Assumethat f

(x) existsandf

(x) > 0for all x. Showthat f (x) cannot benegativefor all x. Hint: Show
that f

(b) = 0for someb andusetheresult of Exercise56inSection4.4.
solution Letf (x) beafunctionsuchthatf

(x) existsandf

(x) > 0for all x. Sincef

(x) > 0, thereisatleastone
point x = b suchthat f

(b) = 0. If not, f

(x) = 0for all x, sof

(x) = 0. Bytheresult of Exercise56inSection4.4,
f (x) ≥ f (b) +f

(b)(x −b). Now, if f

(b) > 0, weﬁndthat f (b) +f

(b)(x −b) > 0whenever
x >
bf

(b) −f (b)
f

(b)
,
aconditionthat must bemet for somex sufﬁciently large. For suchx, f (x) > f (b) + f

(b)(x − b) > 0. Ontheother
hand, if f

(b) < 0, weﬁndthat f (b) +f

(b)(x −b) > 0whenever
x <
bf

(b) −f (b)
f

(b)
.
For suchanx, f (x) > f (b) +f

(b)(x −b) > 0.
4.6 Applied Optimization
Preliminary Questions
1. Theproblemis to ﬁndtheright triangleof perimeter 10whoseareais as largeas possible. What is theconstraint
equationrelatingthebaseb andheight h of thetriangle?
solution Theperimeter of aright triangleisthesumof thelengthsof thebase, theheight andthehypotenuse. If the
basehaslengthb andtheheight ish, thenthelengthof thehypotenuseis
_
b
2
+h
2
andtheperimeter of thetriangleis
P = b +h +
_
b
2
+h
2
. Therequirement that theperimeter be10translatestotheconstraint equation
b +h +
_
b
2
+h
2
= 10.
2. Describeawayof showingthat acontinuousfunctiononanopeninterval (a, b) hasaminimumvalue.
solution If thefunctiontendstoinﬁnity at theendpointsof theinterval, thenthefunctionmust takeonaminimum
valueat acritical point.
3. Istherearectangleof area100of largest perimeter? Explain.
solution No. Evenbyﬁxingtheareaat 100, wecantakeoneof thedimensionsaslargeasweliketherebyallowing
theperimeter tobecomeaslargeaswelike.
Exercises
1. Findthedimensionsx andy of therectangleof maximumareathat canbeformedusing3metersof wire.
(a) What istheconstraint equationrelatingx andy?
(b) Findaformulafor theareaintermsof x alone.
(c) What istheinterval of optimization? Isit openor closed?
(d) Solvetheoptimizationproblem.
solution
(a) Theperimeter of therectangleis3meters, so3= 2x +2y, whichisequivalent toy =
3
2
−x.
(b) Usingpart (a), A = xy = x(
3
2
−x) =
3
2
x −x
2
.
(c) Thisproblemrequiresoptimizationover theclosedinterval [0,
3
2
], sincebothx andy must benon-negative.
(d) A

(x) =
3
2
− 2x = 0, whichyields x =
3
4
andconsequently, y =
3
4
. BecauseA(0) = A(3/2) = 0andA(
3
4
) =
0.5625, themaximumarea0.5625m
2
isachievedwithx = y =
3
4
m.
Wireof length12misdividedintotwopiecesandeachpieceisbent intoasquare. Howshouldthisbedonein
order tominimizethesumof theareasof thetwosquares?
(a) Expressthesumof theareasof thesquaresintermsof thelengthsx andy of thetwopieces.
(b) What istheconstraint equationrelatingx andy?
(c) What istheinterval of optimization? Isit openor closed?
June 9, 2011 LTSV SSM Second Pass
S E C T I ON 4.6 Applied Optimization 221
3. Wireof length12misdividedintotwopiecesandthepiecesarebent intoasquareandacircle. Howshouldthisbe
doneinorder tominimizethesumof their areas?
solution Supposethewireis divided into onepieceof length x mthat is bent into acircleand apieceof length
12−x mthatisbentintoasquare. Becausethecirclehascircumferencex, itfollowsthattheradiusof thecircleisx/2π;
therefore, theareaof thecircleis
π
_
x

_
2
=
x
2

.
Asfor thesquare, becausetheperimeter is12−x, thelengthof eachsideis3−x/4andtheareais(3−x/4)
2
. Then
A(x) =
x
2

+
_
3−
1
4
x
_
2
.
Now
A

(x) =
x

1
2
_
3−
1
4
x
_
= 0
when
x =
12π
4+π
m≈ 5.28m.
BecauseA(0) = 9m
2
, A(12) = 36/π ≈ 11.46m
2
, and
A
_
12π
4+π
_
≈ 5.04m
2
,
weseethat thesumof theareasisminimizedwhenapproximately5.28mof thewireisallottedtothecircle.
Findthepositivenumber x suchthat thesumof x andits reciprocal is as small as possible. Does this problem
requireoptimizationover anopeninterval or aclosedinterval?
5. A ﬂexibletubeof length 4 mis bent into an L-shape. Whereshould thebend bemadeto minimizethedistance
betweenthetwoends?
solution Let x, y > 0belengthsof thesideof theL. Sincex +y = 4or y = 4−x, thedistancebetweentheends
of L ish(x) =
_
x
2
+y
2
=
_
x
2
+(4−x)
2
. Wemayequivalentlyminimizethesquareof thedistance,
f (x) = x
2
+y
2
= x
2
+(4−x)
2
This is easier computationally (whenworkingby hand). Solvef

(x) = 4x − 8 = 0to obtainx = 2m. Nowf (0) =
f (4) = 16, whereasf (2) = 8. Hencethedistancebetweenthetwoendsof theL isminimizedwhenthebendismadeat
themiddleof thewire.
Findthedimensionsof theboxwithsquarebasewith:
(a) Volume12andtheminimal surfacearea.
(b) Surfacearea20andmaximal volume.
7. A rancher will use600mof fencingto buildacorral intheshapeof asemicircleontopof arectangle(Figure9).
Findthedimensionsthat maximizetheareaof thecorral.
FIGURE 9
solution Let x bethewidthof thecorral andthereforethediameter of thesemicircle, andlet y betheheight of the
rectangular section. Thentheperimeter of thecorral canbeexpressedbytheequation2y +x +
π
2
x = 2y +(1+
π
2
)x =
600 mor equivalently, y =
1
2
_
600−(1+
π
2
)x
_
. Since x and y must both be nonnegative, it follows that x must
berestricted to theinterval [0,
600
1+π/2
]. Theareaof thecorral is thesumof theareaof therectangleand semicircle,
A = xy +
π
8
x
2
. Makingthesubstitutionfor y fromtheconstraint equation,
A(x) =
1
2
x
_
600−(1+
π
2
)x
_
+
π
8
x
2
= 300x −
1
2
_
1+
π
2
_
x
2
+
π
8
x
2
.
Now, A

(x) = 300−
_
1+
π
2
_
x +
π
4
x = 0impliesx =
300
_
1+
π
4
_
≈ 168.029746m. WithA(0) = 0m
2
,
A
_
300
1+π/4
_
≈ 25204.5m
2
and A
_
600
1+π/2
_
≈ 21390.8m
2
,
it followsthat thecorral of maximumareahasdimensions
x =
300
1+π/4
m and y =
150
1+π/4
m.
June 9, 2011 LTSV SSM Second Pass
222 C HA P T E R 4 APPLICATIONS OF THE DERIVATIVE
What isthemaximumareaof arectangleinscribedinaright trianglewith5and8asinFigure10. Thesidesof
therectangleareparallel tothelegsof thetriangle.
9. Findthedimensionsof therectangleof maximumareathat canbeinscribedinacircleof radiusr = 4(Figure11).
r
FIGURE 11
solution Placethecenter of thecircleat theoriginwiththesides of therectangle(of lengths 2x > 0and2y > 0)
parallel tothecoordinateaxes. BythePythagoreanTheorem, x
2
+y
2
= r
2
= 16, sothat y =
_
16−x
2
. Thusthearea
of therectangleisA(x) = 2x · 2y = 4x
_
16−x
2
. Toguaranteebothx andy arereal andnonnegative, wemust restrict
x totheinterval [0, 4]. Solve
A

(x) = 4
_
16−x
2

4x
2
_
16−x
2
= 0
for x > 0toobtainx =
4

2
= 2

2. SinceA(0) = A(4) = 0andA(2

2) = 32, therectangleof maximumareahas
dimensions2x = 2y = 4

2.
Findthedimensionsx andy of therectangleinscribedinacircleof radiusr that maximizesthequantityxy
2
.
11. Findthepoint ontheliney = x closest tothepoint (1, 0). Hint: It isequivalent andeasier tominimizethesquare
of thedistance.
solution Withy = x, let’sequivalentlyminimizethesquareof thedistance, f (x) = (x −1)
2
+y
2
= 2x
2
−2x +1,
whichiscomputationallyeasier (whenworkingbyhand). Solvef

(x) = 4x −2= 0toobtainx =
1
2
. Sincef (x) →∞
asx →±∞, (
1
2
,
1
2
) isthepoint ony = x closest to(1, 0).
Findthepoint P ontheparabolay = x
2
closest tothepoint (3, 0) (Figure12).
13. Findthecoordinatesof thepoint onthegraphof y = x +2x
−1
closest totheoriginintheregionx > 0(Figure13).
1 2 3
2
4
6
8
y = x + 2x
−1
y
x
FIGURE 13
solution The distance from the origin to the point (x, x + 2x
−1
) on the graph of y = x + 2x
−1
is d =
_
x
2
+(x +2x
−1
)
2
. Asusual, wewill minimized
2
. Let d
2
= f (x) = x
2
+(x +2x
−1
)
2
. Then
f

(x) = 2x +2(x +2x
−1
)(1−2x
−2
) = 4x −8x
−3
.
Todeterminex, weneedtosolvef

(x) = 0. Multiplyingthroughby x
3
gives4x
4
− 8 = 0, sothat x =
4

2 ≈ 1.189.
Sincef (1.189) ≈ 2.871, thepoint ony = x +2x
−1
that isclosest totheoriginisapproximately(1.189, 2.871).
Problemof Tartaglia(1500–1557) Amongall positivenumbersa, b whosesumis8, ﬁndthosefor whichthe
product of thetwonumbersandtheir differenceislargest.
15. Findtheangleθ that maximizestheareaof theisoscelestrianglewhoselegshavelength (Figure14).
q
FIGURE 14
solution Theareaof thetriangleis
A(θ) =
1
2

2
sinθ,
where0≤ θ ≤ π. Setting
A

(θ) =
1
2

2
cosθ = 0
yieldsθ =
π
2
. SinceA(0) = A(π) = 0andA(
π
2
) =
1
2

2
, theanglethat maximizestheareaof theisoscelestriangleis
θ =
π
2
.
June 9, 2011 LTSV SSM Second Pass
S E C T I ON 4.6 Applied Optimization 223
Arightcircularcone(Figure15)hasvolumeV =
π
3
r
2
h andsurfaceareaisS = πr
_
r
2
+h
2
. Findthedimensions
of theconewithsurfacearea1andmaximal volume.
17. Findtheareaof thelargest isoscelestrianglethat canbeinscribedinacircleof radiusr.
solution Consider thefollowingdiagram:
q
p −q p −q
2q
r
r r
Theareaof theisoscelestriangleis
A(θ) = 2·
1
2
r
2
sin(π −θ) +
1
2
r
2
sin(2θ) = r
2
sinθ +
1
2
r
2
sin(2θ),
where0≤ θ ≤ π. Solve
A

(θ) = r
2
cosθ +r
2
cos(2θ) = 0
toobtainθ =
π
3
, π. SinceA(0) = A(π) = 0andA(
π
3
) =
3

3
4
r
2
, theareaof thelargest isoscelestrianglethat canbe
inscribedinacircleof radiusr is
3

3
4
r
2
.
Findtheradiusandheight of acylindrical canof total surfaceareaA whosevolumeisaslargeaspossible. Does
thereexist acylinder of surfaceareaA andminimal total volume?
19. A poster of area6000cm
2
hasblank marginsof width10cmonthetopandbottomand6cmonthesides. Findthe
dimensionsthat maximizetheprintedarea.
solution Letx bethewidthof theprintedregion, andlety betheheight. Thetotal printedareaisA = xy. Becausethe
total areaof theposter is6000cm
2
, wehavetheconstraint (x +12)(y +20) = 6000, sothat xy +12y +20x +240=
6000, or y =
5760−20x
x+12
. Therefore, A(x) = 20
288x−x
2
x+12
, where0≤ x ≤ 288.
A(0) = A(288) = 0, sowearelookingfor acritical point ontheinterval [0, 288]. SettingA

(x) = 0yields
20
(x +12)(288−2x) −(288x −x
2
)
(x +12)
2
= 0
−x
2
−24x +3456
(x +12)
2
= 0
x
2
+24x −3456= 0
(x −48)(x +72) = 0
Thereforex = 48or x = −72. x = 48istheonlycritical point of A(x) intheinterval [0, 288], soA(48) = 3840isthe
maximumvalueof A(x) intheinterval [0, 288]. Now, y = 20
288−48
48+12
= 80cm, sotheposter withmaximumprintedarea
is48+12= 60cm. wideby80+20= 100cm. tall.
Accordingtopostal regulations, acartonisclassiﬁedas“oversized” if thesumof itsheight andgirth( perimeter
of itsbase) exceeds108in. Findthedimensionsof acartonwithsquarebasethat isnot oversizedandhasmaximum
volume.
21. Kepler’sWineBarrel Problem In his work Nova stereometria doliorum vinariorum (NewSolid Geometry of
a Wine Barrel), published in 1615, astronomer J ohannes Kepler stated and solved the following problem: Find the
dimensionsof thecylinder of largest volumethat canbeinscribedinasphereof radiusR. Hint: Showthat aninscribed
cylinder hasvolume2πx(R
2
−x
2
), wherex isone-half theheight of thecylinder.
solution Placethecenter of thesphereat theorigininthree-dimensional space. Let thecylinder beof radiusy and
half-height x. ThePythagoreanTheoremstates, x
2
+ y
2
= R
2
, so that y
2
= R
2
− x
2
. Thevolumeof thecylinder is
V(x) = πy
2
(2x) = 2π
_
R
2
−x
2
_
x = 2πR
2
x − 2πx
3
. Allowing for degeneratecylinders, wehave0 ≤ x ≤ R.
SolveV

(x) = 2πR
2
− 6πx
2
= 0 for x ≥ 0 to obtain x =
R

3
. SinceV(0) = V(R) = 0, thelargest volumeis
V(
R

3
) =
4
9
π

3R
3
whenx =
R

3
andy =
_
2
3
R.
Find theangleθ that maximizes theareaof thetrapezoid with abaseof length 4 and sides of length 2, as in
Figure16.
23. A landscapearchitect wishes toenclosearectangular gardenof area1,000m
2
ononesideby abrick wall costing
\$90/mandontheother threesidesbyametal fencecosting\$30/m. Whichdimensionsminimizethetotal cost?
solution Let x bethelengthof thebrickwall andy thelengthof anadjacent sidewithx, y > 0. Withxy = 1000or
y =
1000
x
, thetotal cost is
C(x) = 90x +30(x +2y) = 120x +60,000x
−1
.
SolveC

(x) = 120−60,000x
−2
= 0for x > 0toobtainx = 10

5. SinceC(x) →∞asx →0+ andasx →∞, the
minimumcost isC(10

5) = 2400

5≈ \$5366.56whenx = 10

5≈ 22.36mandy = 20

5≈ 44.72m.
Theamount of light reachingapoint at adistancer fromalight sourceA of intensityI
A
isI
A
/r
2
. Supposethat
asecondlight sourceB of intensityI
B
= 4I
A
islocated10mfromA. Findthepoint onthesegment joiningA and
B wherethetotal amount of light isat aminimum.
June 9, 2011 LTSV SSM Second Pass
224 C HA P T E R 4 APPLICATIONS OF THE DERIVATIVE
25. Find themaximumareaof arectangleinscribed in theregion bounded by thegraph of y =
4−x
2+x
and theaxes
(Figure17).
2
4
y =
4− x
2 + x
x
y
FIGURE 17
solution Let s bethewidthof therectangle. Theheight of therectangleish =
4−s
2+s
, sothat theareais
A(s) = s
4−s
2+s
=
4s −s
2
2+s
.
Wearemaximizingontheclosedinterval [0, 4]. It is obvious fromthepictures that A(0) = A(4) = 0, sowelook for
critical pointsof A.
A

(s) =
(2+s)(4−2s) −(4s −s
2
)
(2+s)
2
= −
s
2
+4s −8
(s +2)
2
.
Theonlypoint whereA

(s) doesn’t exist iss = −2whichisn’t under consideration.
SettingA

(s) = 0gives, bythequadraticformula,
s =
−4±

48
2
= −2±2

3.
Of these, only−2+2

3ispositive, sothisisour lonecritical point. A(−2+2

3) ≈ 1.0718> 0. Sinceweareﬁnding
themaximumoveraclosedinterval and−2+2

3istheonlycritical point,themaximumareaisA(−2+2

3) ≈ 1.0718.
Findthemaximumareaof atriangleformedby theaxesandatangent linetothegraphof y = (x + 1)
−2
with
x > 0.
27. Findthemaximumareaof arectanglecircumscribedaroundarectangleof sidesL andH. Hint: Expresstheareain
termsof theangleθ (Figure18).
H
q
L
FIGURE 18
solution PositiontheL ×H rectangleintheﬁrst quadrant of thexy-planewithits“northwest” corner at theorigin.
Letθ betheanglethebaseof thecircumscribedrectanglemakeswiththepositivex-axis, where0≤ θ ≤
π
2
. Thenthearea
of thecircumscribedrectangleisA = LH + 2·
1
2
(H sinθ)(H cosθ) + 2·
1
2
(Lsinθ)(Lcosθ) = LH +
1
2
(L
2
+ H
2
)
sin2θ, whichhas amaximumvalueof LH +
1
2
(L
2
+ H
2
) whenθ =
π
4
becausesin2θ achieves its maximumwhen
θ =
π
4
.
Acontractorisengagedtobuildstepsuptheslopeof ahill thathastheshapeof thegraphof y = x
2
(120−x)/6400
for 0≤ x ≤ 80withx inmeters(Figure19). Whatisthemaximumvertical riseof astair if eachstair hasahorizontal
lengthof one-thirdmeter.
29. Findtheequationof thelinethroughP = (4, 12) suchthat thetriangleboundedbythislineandtheaxesintheﬁrst
quadrant hasminimal area.
solution Let P = (4, 12) be a point in the ﬁrst quadrant and y − 12 = m(x − 4), −∞ < m < 0, be a line
throughP that cutsthepositivex- andy-axes. Theny = L(x) = m(x −4) +12. ThelineL(x) intersectsthey-axisat
H (0, 12−4m) andthex-axisat W
_
4−
12
m
, 0
_
. Hencetheareaof thetriangleis
A(m) =
1
2
(12−4m)
_
4−
12
m
_
= 48−8m−72m
−1
.
Solve A

(m) = 72m
−2
− 8 = 0 for m < 0 to obtain m = −3. Since A → ∞ as m → −∞ or m → 0−, we
concludethat theminimal triangular areais obtainedwhenm = −3. Theequationof thelinethroughP = (4, 12) is
y = −3(x −4) +12= −3x +24.
LetP = (a, b) lieintheﬁrstquadrant. Findtheslopeof thelinethroughP suchthatthetriangleboundedbythis
lineandtheaxesintheﬁrst quadrant hasminimal area. Thenshowthat P isthemidpoint of thehypotenuseof this
triangle.
June 9, 2011 LTSV SSM Second Pass
S E C T I ON 4.6 Applied Optimization 225
31. Archimedes’ Problem A spherical cap(Figure20) of radius r andheight h has volumeV = πh
2
_
r −
1
3
h
_
and
surfaceareaS = 2πrh. Provethat thehemisphereenclosesthelargest volumeamongall spherical capsof ﬁxedsurface
areaS.
r
h
FIGURE 20
solution Consider all spherical capsof ﬁxedsurfaceareaS. BecauseS = 2πrh, it followsthat
r =
S
2πh
and
V(h) = πh
2
_
S
2πh

1
3
h
_
=
S
2
h −
π
3
h
3
.
Now
V

(h) =
S
2
−πh
2
= 0
when
h
2
=
S

or h =
S
2πh
= r.
Hence, thehemisphereenclosesthelargest volumeamongall spherical capsof ﬁxedsurfaceareaS.
Find the isosceles triangle of smallest area (Figure 21) that circumscribes a circle of radius 1 (fromThomas
Simpson’sThe Doctrine and Application of Fluxions, acalculustext that appearedin1750).
33. A boxof volume72m
3
withsquarebottomandnotopisconstructedout of twodifferent materials. Thecost of the
bottomis\$40/m
2
andthecost of thesidesis\$30/m
2
. Findthedimensionsof theboxthat minimizetotal cost.
solution Let s denotethelengthof thesideof thesquarebottomof theboxandh denotetheheight of thebox. Then
V = s
2
h = 72 or h =
72
s
2
.
Thecost of theboxis
C = 40s
2
+120sh = 40s
2
+
8640
s
,
so
C

(s) = 80s −
8640
s
2
= 0
whens = 3
3

4mandh = 2
3

4m. BecauseC → ∞as s → 0− andas s → ∞, weconcludethat thecritical point
givestheminimumcost.
Findthedimensionsof acylinder of volume1m
3
of minimal costif thetopandbottomaremadeof material that
coststwiceasmuchasthematerial for theside.
35. Your task is to design a rectangular industrial warehouse consisting of three separate spaces of equal size as in
Figure22. Thewall materialscost \$500per linear meter andyour companyallocates\$2,400,000for theproject.
(a) Whichdimensionsmaximizetheareaof thewarehouse?
(b) What istheareaof eachcompartment inthiscase?
FIGURE 22
solution Letonecompartmenthavelengthx andwidthy.Thentotal lengthof thewall of thewarehouseisP = 4x +6y
andtheconstraint equationiscost = 2,400,000= 500(4x +6y), whichgivesy = 800−
2
3
x.
(a) AreaisgivenbyA = 3xy = 3x
_
800−
2
3
x
_
= 2400x −2x
2
, where0≤ x ≤ 1200. ThenA

(x) = 2400−4x = 0
yieldsx = 600andconsequentlyy = 400. SinceA(0) = A(1200) = 0andA(600) = 720, 000, theareaof thewarehouse
ismaximizedwheneachcompartment haslengthof 600 mandwidthof 400 m.
(b) Theareaof onecompartment is600· 400= 240, 000squaremeters.
June 9, 2011 LTSV SSM Second Pass
226 C HA P T E R 4 APPLICATIONS OF THE DERIVATIVE
Suppose, inthepreviousexercise, that thewarehouseconsistsof n separatespacesof equal size. Findaformula
intermsof n for themaximumpossibleareaof thewarehouse.
37. Accordingtoamodel developedbyeconomistsE. HeadyandJ . Pesek, if fertilizer madefromN poundsof nitrogen
andP poundsof phosphateisusedonanacreof farmland, thentheyieldof corn(inbushelsper acre) is
Y = 7.5+0.6N +0.7P −0.001N
2
−0.002P
2
+0.001NP
A farmer intendstospend\$30per acreonfertilizer. If nitrogencosts25cents/lbandphosphatecosts20cents/lb, which
combinationof N andL producesthehighest yieldof corn?
solution Thefarmer’sbudget for fertilizer is\$30per acre, sowehavetheconstraint equation
0.25N +0.2P = 30 or P = 150−1.25N
Substitutingfor P intheequationfor Y, weﬁnd
Y(N) = 7.5+0.6N +0.7(150−1.25N) −0.001N
2
−0.002(150−1.25N)
2
+0.001N(150−1.25N)
= 67.5+0.625N −0.005375N
2
BothN andP must benonnegative. SinceP = 150−1.25N ≥ 0, werequirethat 0≤ N ≤ 120. Next,
dY
dN
= 0.625−0.01075N = 0 ⇒ N =
0.625
0.01075
≈ 58.14pounds.
Now, Y(0) = 67.5, Y(120) = 65.1andY(58.14) ≈ 85.67, sothemaximumyieldof cornoccursfor N ≈ 58.14pounds
andP ≈ 77.33pounds.
Experimentsshowthat thequantitiesx of cornandy of soybeanrequiredtoproduceahogof weight Q satisfy
Q = 0.5x
1/2
y
1/4
. Theunit of x, y, andQ isthecwt, anagricultural unit equal to100lbs. Findthevaluesof x and
y that minimizethecost of ahogof weight Q = 2.5cwt if corncosts\$3/cwt andsoycosts\$7/cwt.
39. All unitsina100-unit apartment buildingarerentedout whenthemonthly rent isset at r = \$900/month. Suppose
thatoneunitbecomesvacantwitheach\$10increaseinrentandthateachoccupiedunitcosts\$80/monthinmaintenance.
Whichrent r maximizesmonthlyproﬁt?
solution Let n denotethenumber of \$10increasesinrent. Thenthemonthlyproﬁt isgivenby
P(n) = (100−n)(900+10n −80) = 82000+180n −10n
2
,
and
P

(n) = 180−20n = 0
whenn = 9. Weknowthisresultsinmaximumproﬁt becausethisgivesthelocationof vertex of adownwardopening
parabola. Thus, monthlyproﬁt ismaximizedwitharent of \$990.
An8-billion-bushel corncropbringsapriceof \$2.40/bu. A commoditybroker usestheruleof thumb: If thecrop
isreducedbyx percent, thenthepriceincreasesby10x cents. Whichcropsizeresultsinmaximumrevenueandwhat
isthepriceper bu? Hint: Revenueisequal topricetimescropsize.
41. Themonthly output of aSpanishlight bulbfactory isP = 2LK
2
(inmillions), whereL isthecost of labor andK
isthecostof equipment(inmillionsof euros). Thecompanyneedstoproduce1.7millionunitsper month. Whichvalues
of L andK wouldminimizethetotal cost L +K?
solution SinceP = 1.7andP = 2LK
2
, wehaveL =
0.85
K
2
. Accordingly, thecost of productionis
C(K) = L +K = K +
0.85
K
2
.
SolveC

(K) = 1−
1.7
K
3
for K ≥ 0toobtainK =
3

1.7. SinceC(K) →∞asK →0+ andasK →∞, theminimum
cost of production is achieved for K =
3

1.7 ≈ 1.2 and L = 0.6. Thecompany should invest 1.2 million euros in
equipment and600, 000eurosinlabor.
Therectangular plot inFigure23hassize100m× 200m. PipeistobelaidfromA toapoint P onsideBC and
fromtheretoC. Thecost of layingpipealongthesideof theplot is \$45/mandthecost throughtheplot is \$80/m
(sinceit isunderground).
(a) Letf (x) bethetotal cost, wherex isthedistancefromP toB. Determinef (x), butnotethatf isdiscontinuous
at x = 0(whenx = 0, thecost of theentirepipeis\$45/ft).
(b) What isthemost economical waytolaythepipe?What if thecost alongthesidesis\$65/m?
43. Brandonis ononesideof ariver that is 50mwideandwants toreachapoint 200mdownstreamontheopposite
sideasquickly aspossibleby swimmingdiagonally acrosstheriver andthenrunningtherest of theway. Findthebest
routeif Brandoncanswimat 1.5m/sandrunat 4m/s.
solution Let lengthsbeinmeters, timesinseconds, andspeedsinm/s. Supposethat Brandonswimsdiagonally to
apoint locatedx metersdownstreamontheoppositeside. ThenBrandonthenswimsadistance
_
x
2
+50
2
andrunsa
distance200−x. Thetotal timeof thetripis
f (x) =
_
x
2
+2500
1.5
+
200−x
4
, 0≤ x ≤ 200.
Solve
f

(x) =
2x
3
_
x
2
+2500

1
4
= 0
toobtainx = 30
5
11
≈ 20.2andf (20.2) ≈ 80.9. Sincef (0) ≈ 83.3andf (200) ≈ 137.4, weconcludethat theminimal
timeis80.9s. ThisoccurswhenBrandonswimsdiagonallytoapoint located20.2mdownstreamandthenrunstherest
of theway.
June 9, 2011 LTSV SSM Second Pass
S E C T I ON 4.6 Applied Optimization 227
Snell’sLaw Whenalight beamtravels fromapoint A aboveaswimmingpool toapoint B belowthewater
(Figure24), it choosesthepaththat takestheleast time. Let v
1
bethevelocity of light inair andv
2
thevelocity in
water (it isknownthat v
1
> v
2
). ProveSnell’sLawof Refraction:
sinθ
1
v
1
=
sinθ
2
v
2
In Exercises 45–47, a box (with no top) is to be constructed from a piece of cardboard of sides A and B by cutting out
squares of length h from the corners and folding up the sides (Figure 26).
h
A
B
FIGURE 26
45. Findthevalueof h that maximizesthevolumeof thebox if A = 15andB = 24. What arethedimensionsof this
box?
solution Oncethesideshavebeenfoldedup, thebaseof theboxwill havedimensions(A−2h) ×(B −2h) andthe
height of theboxwill beh. Thus
V(h) = h(A −2h)(B −2h) = 4h
3
−2(A +B)h
2
+ABh.
WhenA = 15andB = 24, thisgives
V(h) = 4h
3
−78h
2
+360h,
andweneedtomaximizeover 0≤ h ≤
15
2
. Now,
V

(h) = 12h
2
−156h +360= 0
yieldsh = 3andh = 10. Becauseh = 10isnotinthedomainof theproblemandV(0) = V(15/2) = 0andV(3) = 486,
volumeismaximizedwhenh = 3. Thecorrespondingdimensionsare9×18×3.
Vascular Branching A small bloodvessel of radiusr branchesoff at anangleθ fromalarger vessel of radius
R to supply blood along apath fromA to B. According to Poiseuille’s Law, thetotal resistanceto blood ﬂowis
proportional to
T =
_
a −b cotθ
R
4
+
b cscθ
r
4
_
wherea andb areasinFigure25. Showthat thetotal resistanceisminimizedwhencosθ = (r/R)
4
.
47. Whichvaluesof A andB maximizethevolumeof theboxif h = 10cmandAB = 900cm.
solution Withh = 10andAB = 900(whichmeansthat B = 900/A), thevolumeof theboxis
V(A) = 10(A −20)
_
900
A
−20
_
= 13,000−200A −
180,000
A
,
where20≤ A ≤ 45. Now, solving
V

(A) = −200+
180,000
A
2
= 0
yields A = 30. BecauseV(20) = V(45) = 0andV(30) = 1000cm
3
, maximumvolumeis achievedwithA = B =
30cm.
Givenn numbersx
1
, . . . , x
n
, ﬁndthevalueof x minimizingthesumof thesquares:
(x −x
1
)
2
+(x −x
2
)
2
+· · · +(x −x
n
)
2
First solvefor n = 2, 3andthentryit for arbitraryn.
49. A billboardof height b ismountedonthesideof abuildingwithitsbottomedgeat adistanceh fromthestreet asin
Figure27. At what distancex shouldanobserver standfromthewall tomaximizetheangleof observationθ?
h
b
x
P
q θ
ψ
ψ
P
A
R
B
C
Q
FIGURE 27
solution Fromtheupper diagraminFigure27andtheadditionformulafor thecotangent function, weseethat
cotθ =
1+
x
b+h
x
h
x
h

x
b+h
=
x
2
+h(b +h)
bx
,
whereb andh areconstant. Now, differentiatewithrespect tox andsolve
−csc
2
θ

dx
=
x
2
−h(b +h)
bx
2
= 0
toobtainx =
_
bh +h
2
. Sincethisistheonlycritical point, andsinceθ →0asx →0+ andθ →0asx →∞, θ(x)
reachesitsmaximumat x =
_
bh +h
2
.
June 9, 2011 LTSV SSM Second Pass
228 C HA P T E R 4 APPLICATIONS OF THE DERIVATIVE
SolveExercise49againusinggeometry rather thancalculus. ThereisauniquecirclepassingthroughpointsB
andC whichistangent tothestreet. Let R bethepoint of tangency. Notethat thetwoangleslabeledψ inFigure27
areequal becausetheysubtendequal arcsonthecircle.
(a) Showthat themaximumvalueof θ isθ = ψ. Hint: Showthat ψ = θ +

PBA whereA istheintersectionof
thecirclewithPC.
(b) Provethat thisagreeswiththeanswer toExercise49.
(c) Showthat

QRB =

RCQfor themaximal angleψ.
51. Optimal DeliverySchedule A gas station sells Q gallons of gasolineper year, which is delivered N times per
year inequal shipments of Q/N gallons. Thecost of eachdelivery is d dollars andtheyearly storagecosts aresQT ,
whereT isthelengthof time(afractionof ayear) betweenshipmentsands isaconstant. Showthat costsareminimized
for N =

sQ/d. (Hint: T = 1/N.) Find theoptimal number of deliveries if Q = 2 million gal, d = \$8000, and
s = 30cents/gal-yr. Your answer shouldbeawholenumber, socomparecostsfor thetwointeger valuesof N nearestthe
optimal value.
solution ThereareN shipmentsper year, sothetimeinterval betweenshipmentsisT = 1/N years. Hence, thetotal
storagecostsper year aresQ/N. TheyearlydeliverycostsaredN andthetotal costsisC(N) = dN +sQ/N. Solving,
C

(N) = d −
sQ
N
2
= 0
for N yieldsN =

sQ/d. For thespeciﬁccaseQ = 2,000,000, d = 8000ands = 0.30,
N =
_
0.30(2,000,000)
8000
= 8.66.
WithC(8) = \$139,000andC(9) = \$138,667, theoptimal number of deliveriesper year isN = 9.
VictorKlee’sEndpointMaximumProblemGiven40metersof straightfence, yourgoal istobuildarectangular
enclosureusing80additional metersof fencethatencompassesthegreatestarea. LetA(x) betheareaof theenclosure,
withx asinFigure28.
(a) Findthemaximumvalueof A(x).
(b) Whichinterval of x valuesisrelevant toour problem? Findthemaximumvalueof A(x) onthisinterval.
53. Let(a, b) beaﬁxedpointintheﬁrstquadrantandletS(d) bethesumof thedistancesfrom(d, 0) tothepoints(0, 0),
(a, b), and(a, −b).
(a) Findthevalueof d for whichS(d) isminimal. Theanswer dependsonwhether b <

3a or b ≥

3a. Hint: Show
that d = 0whenb ≥

3a.
(b) Let a = 1. Plot S(d) for b = 0.5,

3, 3anddescribethepositionof theminimum.
solution
(a) If d < 0, then thedistancefrom(d, 0) to theother threepoints can all bereduced by increasing thevalueof d.
Similarly, if d > a, thenthedistancefrom(d, 0) totheother threepointscanall bereducedby decreasingthevalueof
d. It followsthat theminimumof S(d) must occur for 0≤ d ≤ a. Restrictingattentiontothisinterval, weﬁnd
S(d) = d +2
_
(d −a)
2
+b
2
.
Solving
S

(d) = 1+
2(d −a)
_
(d −a)
2
+b
2
= 0
yieldsthecritical point d = a −b/

3. If b <

3a, thend = a −b/

3> 0andtheminimumoccursatthisvalueof d.
Ontheother hand, if b ≥

3a, thentheminimumoccursat theendpoint d = 0.
(b) Leta = 1. Plotsof S(d) for b = 0.5, b =

3andb = 3areshownbelow. For b = 0.5, theresultsof (a) indicatethe
minimumshouldoccur for d = 1−0.5/

3≈ 0.711, andthisisconﬁrmedintheplot. For bothb =

3andb = 3, the
resultsof (a) indicatethat theminimumshouldoccur at d = 0, andbothof theseconclusionsareconﬁrmedintheplots.
1.6
x
0 0.2 0.4 0.6 0.8
b = 0.5
1
y
1.5
1.9
1.8
1.7
2
2.1
x
0 0.2 0.4 0.6 0.8 1
y
6.6
6.4
6.8
b = 3
4
x
0 0.2 0.4 0.6 0.8 1
y
4.2
4.1
4.3
4.4
b = 3
TheforceF (inNewtons)requiredtomoveaboxof massmkginmotionbypullingonanattachedrope(Figure29)
is
F(θ) =
f mg
cosθ +f sinθ
whereθ istheanglebetweentheropeandthehorizontal, f isthecoefﬁcientof staticfriction, andg = 9.8m/s
2
. Find
theangleθ thatminimizestherequiredforceF, assumingf = 0.4. Hint: Findthemaximumvalueof cosθ +f sinθ.
55. Inthesettingof Exercise54, showthat for anyf theminimal forcerequiredisproportional to1/
_
1+f
2
.
solution WeminimizeF(θ) by ﬁndingthemaximumvalueg(θ) = cosθ + f sinθ. Theangleθ isrestrictedtothe
interval [0,
π
2
]. Wesolvefor thecritical points:
g

(θ) = −sinθ +f cosθ = 0
Weobtain
f cosθ = sinθ ⇒ tanθ = f
Fromtheﬁgurebelowweﬁndthat cosθ = 1/
_
1+f
2
andsinθ = f/
_
1+f
2
. Hence
g(θ) =
1
f
+
f
2
_
1+f
2
=
1+f
2
_
1+f
2
=
_
1+f
2
June 9, 2011 LTSV SSM Second Pass
S E C T I ON 4.6 Applied Optimization 229
Thevaluesat theendpointsare
g(0) = 1, g
_
π
2
_
= f
Bothof thesevaluesarelessthan
_
1+f
2
. Thereforethemaximumvalueof g(θ) is
_
1+f
2
andtheminimumvalue
of F(θ) is
F =
f mg
g(θ)
=
f mg
_
1+f
2
f
α
1
√1 + f
2
BirdMigration Ornithologistshavefoundthat thepower (injoulesper second) consumedbyacertainpigeon
ﬂyingat velocityv m/sisdescribedwell bythefunctionP(v) = 17v
−1
+10
−3
v
3
J /s. Assumethat thepigeoncan
store5×10
4
J of usableenergyasbodyfat.
(a) Showthat at velocity v, apigeoncanﬂy atotal distanceof D(v) = (5× 10
4
)v/P(v) if it usesall of itsstored
energy.
(b) Findthevelocityv
p
that minimizes P(v).
(c) Migratingbirdsaresmart enoughtoﬂy at thevelocity that maximizesdistancetraveledrather thanminimizes
power consumption. Showthat thevelocity v
d
whichmaximizesD(v) satisﬁesP

(v
d
) = P(v
d
)/v
d
. Showthat v
d
isobtainedgraphicallyasthevelocitycoordinateof thepoint wherealinethroughtheoriginistangent tothegraph
of P(v) (Figure30).
(d) Findv
d
andthemaximumdistanceD(v
d
).
57. Theproblemistoputa“roof”of sides onanatticroomof heighth andwidthb. Findthesmallestlengths for which
thisispossibleif b = 27andh = 8(Figure31).
s
h
b
FIGURE 31
solution Consider theright triangleformed by theright half of therectangleand its “roof”. This trianglehas hy-
potenuses.
h
b/2 x
s y
Asshown, let y betheheight of theroof, andlet x bethedistancefromtheright baseof therectangletothebaseof the
roof. Bysimilar trianglesappliedtothesmaller right trianglesat thetopandright of thelarger triangle, weget:
y −8
27/2
=
8
x
or y =
108
x
+8.
s, y, andx arerelatedbythePythagoreanTheorem:
s
2
=
_
27
2
+x
_
2
+y
2
=
_
27
2
+x
_
2
+
_
108
x
+8
_
2
.
Sinces > 0, s
2
isleastwhenever s isleast, sowecanminimizes
2
insteadof s. Settingthederivativeequal tozeroyields
2
_
27
2
+x
_
+2
_
108
x
+8
__

108
x
2
_
= 0
2
_
27
2
+x
_
+2
8
x
_
27
2
+x
__

108
x
2
_
= 0
2
_
27
2
+x
__
1−
864
x
3
_
= 0
Thezerosarex = −
27
2
(irrelevant) andx = 6
3

4. Sincethisistheonlycritical point of s withx > 0, andsinces →∞
asx →0ands →∞asx →∞, thisisthepoint wheres attainsitsminimum. For thisvalueof x,
s
2
=
_
27
2
+6
3

4
_
2
+
_
9
3

2+8
_
2
≈ 904.13,
sothesmallest roof lengthis
s ≈ 30.07.
June 9, 2011 LTSV SSM Second Pass
230 C HA P T E R 4 APPLICATIONS OF THE DERIVATIVE
RedoExercise57for arbitraryb andh.
59. Findthemaximumlengthof apolethatcanbecarriedhorizontallyaroundacornerjoiningcorridorsof widthsa = 24
andb = 3(Figure32).
a
b
FIGURE 32
solution Inorder to ﬁndthelengthof thelongest polethat canbecarriedaroundthecorridor, wehaveto ﬁndthe
shortest lengthfromtheleft wall tothetopwall touchingthecorner of theinsidewall. Anypolethat doesnot ﬁt inthis
shortest spacecannot becarriedaroundthecorner, soanexact ﬁt representsthelongest possiblepole.
Let θ betheanglebetweenthepoleandahorizontal linetotheright. Let c
1
bethelengthof poleinthecorridor of
width24andlet c
2
bethelengthof poleinthecorridor of width3. Bythedeﬁnitionsof sineandcosine,
3
c
2
= sinθ and
24
c
1
= cosθ,
sothat c
1
=
24
cosθ
, c
2
=
3
sinθ
. What must beminimizedisthetotal length, givenby
f (θ) =
24
cosθ
+
3
sinθ
.
Settingf

(θ) = 0yields
24sinθ
cos
2
θ

3cosθ
sin
2
θ
= 0
24sinθ
cos
2
θ
=
3cosθ
sin
2
θ
24sin
3
θ = 3cos
3
θ
Asθ <
π
2
(thepoleisbeingturnedaroundacorner, after all), wecandividebothsidesby cos
3
θ, gettingtan
3
θ =
1
8
.
Thisimpliesthat tanθ =
1
2
(tanθ > 0astheangleisacute).
Sincef (θ) → ∞ as θ → 0+ andas θ →
π
2
−, wecantell that theminimum is attainedat θ
0
wheretanθ
0
=
1
2
.
Because
tanθ
0
=
opposite
adjacent
=
1
2
,
wedrawatrianglewithoppositeside1andadjacent side2. ByPythagoras, c =

5, so
sinθ
0
=
1

5
and cosθ
0
=
2

5
.
Fromthis, weget
f (θ
0
) =
24
cosθ
0
+
3
sinθ
0
=
24
2

5+3

5= 15

5.
RedoExercise59for arbitrarywidthsa andb.
61. Findtheminimumlength of abeamthatcanclear afenceof heighth andtouchawall locatedb ftbehindthefence
(Figure33).
b x
h
FIGURE 33
June 9, 2011 LTSV SSM Second Pass
S E C T I ON 4.6 Applied Optimization 231
solution Let y betheheight of thepoint wherethebeamtouchesthewall infeet. Bysimilar triangles,
y −h
b
=
h
x
or y =
bh
x
+h
andbyPythagoras:

2
= (b +x)
2
+
_
bh
x
+h
_
2
.
Wecanminimize
2
rather than, sosettingthederivativeequal tozerogives:
2(b +x) +2
_
bh
x
+h
__

bh
x
2
_
= 2(b +x)
_
1−
h
2
b
x
3
_
= 0.
Thezeroesareb = −x (irrelevant) andx =
3

h
2
b. Since
2
→∞asx →0+ andasx →∞, x =
3

h
2
b corresponds
toaminimumfor
2
. For thisvalueof x, wehave

2
= (b +h
2/3
b
1/3
)
2
+(h +h
1/3
b
2/3
)
2
= b
2/3
(b
2/3
+h
2/3
)
2
+h
2/3
(h
2/3
+b
2/3
)
2
= (b
2/3
+h
2/3
)
3
andso
= (b
2/3
+h
2/3
)
3/2
.
A beamthat clears afenceof height h feet and touches awall b feet behind thefencemust havelength at least =
(b
2/3
+h
2/3
)
3/2
ft.
Whichvalueof h maximizesthevolumeof theboxif A = B?
63. A basketball player stands d feet fromthebasket. Let h andα beas inFigure34. Usingphysics, onecan
showthat if theplayer releases theball at anangleθ, thentheinitial velocity requiredtomaketheball gothroughthe
basket satisﬁes
v
2
=
16d
cos
2
θ(tanθ −tanα)
(a) Explainwhythisformulaismeaningful onlyfor α < θ <
π
2
. Whydoesv approachinﬁnityat theendpointsof this
interval?
(b) Takeα =
π
6
andplot v
2
asafunctionof θ for
π
6
< θ <
π
2
. Verifythat theminimumoccursat θ =
π
3
.
(c) Set F(θ) = cos
2
θ(tanθ −tanα). Explainwhyv isminimizedfor θ suchthat F(θ) ismaximized.
(d) Verify that F

(θ) = cos(α − 2θ) secα (you will need to usetheaddition formulafor cosine) and show that the
maximumvalueof F(θ) on
_
α,
π
2
_
occursat θ
0
=
α
2
+
π
4
.
(e) For agivenα, theoptimal anglefor shootingthebasket isθ
0
becauseit minimizesv
2
andthereforeminimizesthe
energyrequiredtomaketheshot(energyisproportional tov
2
). Showthatthevelocityv
opt
attheoptimal angleθ
0
satisﬁes
v
2
opt
=
32d cosα
1−sinα
=
32d
2
−h +
_
d
2
+h
2
(f) Showwithagraphthatfor ﬁxedd (say, d = 15ft, thedistanceof afreethrow), v
2
opt
isanincreasingfunction
of h. Usethistoexplainwhytaller playershaveanadvantageandwhyit canhelptojumpwhileshooting.
q
a
h
d
FIGURE 34
June 9, 2011 LTSV SSM Second Pass
232 C HA P T E R 4 APPLICATIONS OF THE DERIVATIVE
solution
(a) α = 0correspondstoshootingtheball directlyatthebasketwhileα = π/2correspondstoshootingtheball directly
upward. Inneither caseisit possiblefor theball togointothebasket.
If theangleα is extremely closeto0, theball is shot almost directly at thebasket, sothat it must belaunchedwith
great speed, asit canonlyfall anextremelyshort distanceonthewaytothebasket.
Ontheother hand, if theangleα is extremely closetoπ/2, theball is launchedalmost vertically. This requires the
ball totravel agreat distanceupwardinorder totravel thehorizontal distance. Ineither oneof thesecases, theball hasto
travel at anenormousspeed.
(b)
π
6
π
4
π
3

12
π
2
Theminimumclearlyoccurswhereθ = π/3.
(c) If F(θ) = cos
2
θ (tanθ −tanα),
v
2
=
16d
cos
2
θ (tanθ −tanα)
=
16d
F(θ)
.
Sinceα ≤ θ, F(θ) > 0, hencev
2
issmallest whenever F(θ) isgreatest.
(d) F

(θ) = −2sinθ cosθ (tanθ −tanα) + cos
2
θ
_
sec
2
θ
_
= −2sinθ cosθ tanθ + 2sinθ cosθ tanα + 1. Wewill
applyall thedoubleangleformulas:
cos(2θ) = cos
2
θ −sin
2
θ = 1−2sin
2
θ; sin2θ = 2sinθ cosθ,
getting:
F

(θ) = 2sinθ cosθ tanα −2sinθ cosθ tanθ +1
= 2sinθ cosθ
sinα
cosα
−2sinθ cosθ
sinθ
cosθ
+1
= secα
_
−2sin
2
θ cosα +2sinθ cosθ sinα +cosα
_
= secα
_
cosα
_
1−2sin
2
θ
_
+sinα (2sinθ cosθ)
_
= secα (cosα(cos2θ) +sinα(sin2θ))
= secα cos(α −2θ)
A critical point of F(θ) occurswherecos(α −2θ) = 0, sothat α −2θ = −
π
2
(negativebecause2θ > θ > α), andthis
givesusθ = α/2+π/4. TheminimumvalueF(θ
0
) takesplaceat θ
0
= α/2+π/4.
(e) Pluginθ
0
= α/2+π/4. Toﬁndv
2
opt
wemust simplify
cos
2
θ
0
(tanθ
0
−tanα) =
cosθ
0
(sinθ
0
cosα −cosθ
0
sinα)
cosα
Bytheadditionlawfor sine:
sinθ
0
cosα −cosθ
0
sinα = sin(θ
0
−α) = sin(−α/2+π/4)
andso
cosθ
0
(sinθ
0
cosα −cosθ
0
sinα) = cos(α/2+π/4) sin(−α/2+π/4)
Nowusetheidentity(that followsfromtheadditionlaw):
sinx cosy =
1
2
(sin(x +y) +sin(x −y))
toget
cos(α/2+π/4) sin(−α/2+π/4) = (1/2)(1−sinα)
June 9, 2011 LTSV SSM Second Pass
S E C T I ON 4.6 Applied Optimization 233
Soweﬁnallyget
cos
2
θ
0
(tanθ
0
−tanα) =
(1/2)(1−sinα)
cosα
andtherefore
v
2
opt
=
32d cosα
1−sinα
asclaimed. FromFigure34weseethat
cosα =
d
_
d
2
+h
2
and sinα =
h
_
d
2
+h
2
.
Substitutingthesevaluesintotheexpressionfor v
2
opt
yields
v
2
opt
=
32d
2
−h +
_
d
2
+h
2
.
(f) A sketchof thegraphof v
2
opt
versus h for d = 15feet is givenbelow: v
2
opt
increases withrespect tobasket height
relativetotheshooter. Thisshowsthat theminimumvelocity requiredtolaunchtheball tothebasket dropsasshooter
heightincreases. Thisshowsoneof thewaysheightisanadvantageinfreethrows; ataller shooter neednotshoottheball
ashardtoreachthebasket.
100
200
300
400
500
600
4 5 0 3 2 1
ThreetownsA, B, andC aretobejoinedby anundergroundﬁber cableasillustratedinFigure35(A). Assume
that C is locateddirectly belowthemidpoint of AB. Findthejunctionpoint P that minimizes thetotal amount of
cableused.
(a) FirstshowthatP mustliedirectlyaboveC. Hint: Usetheresultof Example6toshowthatif thejunctionisplaced
at point QinFigure35(B), thenwecanreducethecablelengthbymovingQhorizontallyover tothepoint P lying
aboveC.
(b) Withx asinFigure35(A), let f (x) bethetotal lengthof cableused. Showthat f (x) hasauniquecritical point
c. Computec andshowthat 0≤ c ≤ L if andonlyif D ≤ 2

3L.
(c) Findtheminimumof f (x) on[0, L] intwocases: D = 2, L = 4andD = 8, L = 2.
Further Insights and Challenges
65. TomandAli drivealong ahighway represented by thegraph of f (x) in Figure36. During thetrip, Ali views a
billboardrepresentedbythesegmentBC alongthey-axis. LetQbethey-interceptof thetangentlinetoy = f (x). Show
thatθ ismaximizedatthevalueof x for whichtheangles

QPB and

QCP areequal. ThisgeneralizesExercise50(c)
(whichcorrespondstothecasef (x) = 0). Hints:
(a) Showthat dθ/dx isequal to
(b −c) ·
(x
2
+(xf

(x))
2
) −(b −(f (x) −xf

(x)))(c −(f (x) −xf

(x)))
(x
2
+(b −f (x))
2
)(x
2
+(c −f (x))
2
)
(b) Showthat they-coordinateof Qisf (x) −xf

(x).
(c) Showthat theconditiondθ/dx = 0isequivalent to
PQ
2
= BQ· CQ
(d) Concludethat QPB andQCP aresimilar triangles.
x
x
y
billboard
highway
P = (x, f (x))
y = f (x)
B = (0, b)
C = (0, c)
FIGURE 36
solution
(a) Fromtheﬁgure, weseethat
θ(x) = tan
−1
c −f (x)
x
−tan
−1
b −f (x)
x
.
June 9, 2011 LTSV SSM Second Pass
234 C HA P T E R 4 APPLICATIONS OF THE DERIVATIVE
Then
θ

(x) =
b −(f (x) −xf

(x))
x
2
+(b −f (x))
2

c −(f (x) −xf

(x))
x
2
+(c −f (x))
2
= (b −c)
x
2
−bc +(b +c)(f (x) −xf

(x)) −(f (x))
2
+2xf (x)f

(x)
(x
2
+(b −f (x))
2
)(x
2
+(c −f (x))
2
)
= (b −c)
(x
2
+(xf

(x))
2
−(bc −(b +c)(f (x) −xf

(x)) +(f (x) −xf

(x))
2
)
(x
2
+(b −f (x))
2
)(x
2
+(c −f (x))
2
)
= (b −c)
(x
2
+(xf

(x))
2
−(b −(f (x) −xf

(x)))(c −(f (x) −xf

(x)))
(x
2
+(b −f (x))
2
)(x
2
+(c −f (x))
2
)
.
(b) Thepoint Qisthey-intercept of thelinetangent tothegraphof f (x) at point P. Theequationof thistangent lineis
Y −f (x) = f

(x)(X −x).
They-coordinateof Qisthenf (x) −xf

(x).
(c) Fromtheﬁgure, weseethat
BQ = b −(f (x) −xf

(x)),
CQ = c −(f (x) −xf

(x))
and
PQ =
_
x
2
+(f (x) −(f (x) −xf

(x)))
2
=
_
x
2
+(xf

(x))
2
.
Comparingtheseexpressionswiththenumerator of dθ/dx, it followsthat

dx
= 0isequivalent to
PQ
2
= BQ· CQ.
(d) TheequationPQ
2
= BQ· CQisequivalent to
PQ
BQ
=
CQ
PQ
.
Inother words, thesidesCQandPQfromthetriangleQCP areproportional inlengthtothesidesPQandBQfrom
thetriangleQPB. As

PQB =

CQP, it followsthat trianglesQCP andQPB aresimilar.
SeismicProspecting Exercises 66–68 are concerned with determining the thickness d of a layer of soil that lies on top
of a rock formation. Geologists send two sound pulses from point A to point D separated by a distance s. The ﬁrst pulse
travels directly from A to D along the surface of the earth. The second pulse travels down to the rock formation, then
along its surface, and then back up to D (path ABCD), as in Figure 37. The pulse travels with velocity v
1
in the soil and
v
2
in the rock.
A
B C
s D
Soil
Rock
q q d
FIGURE 37
(a) Showthat thetimerequiredfor theﬁrst pulsetotravel fromA toD ist
1
= s/v
1
.
(b) Showthat thetimerequiredfor thesecondpulseis
t
2
=
2d
v
1
secθ +
s −2d tanθ
v
2
providedthat
tanθ ≤
s
2d
(Note: If thisinequalityisnot satisﬁed, thenpoint B doesnot lietotheleft of C.)
(c) Showthat t
2
isminimizedwhensinθ = v
1
/v
2
.
67. Inthisexercise, assumethat v
2
/v
1

_
1+4(d/s)
2
.
(a) Showthat inequality(2) holdsif sinθ = v
1
/v
2
.
(b) Showthat theminimal timefor thesecondpulseis
t
2
=
2d
v
1
(1−k
2
)
1/2
+
s
v
2
wherek = v
1
/v
2
.
(c) Concludethat
t
2
t
1
=
2d(1−k
2
)
1/2
s
+k.
June 9, 2011 LTSV SSM Second Pass
S E C T I ON 4.6 Applied Optimization 235
solution
(a) If sinθ =
v
1
v
2
, then
tanθ =
v
1
_
v
2
2
−v
2
1
=
1
_
_
v
2
v
1
_
2
−1
.
Because
v
2
v
1

_
1+4(
d
s
)
2
, it followsthat
_
_
v
2
v
1
_
2
−1≥
_
1+4
_
d
s
_
2
−1=
2d
s
.
Hence, tanθ ≤
s
2d
asrequired.
(b) For thetime-minimizingchoiceof θ, wehavesinθ =
v
1
v
2
fromwhichsecθ =
v
2
_
v
2
2
−v
2
1
andtanθ =
v
1
_
v
2
2
−v
2
1
.
Thus
t
2
=
2d
v
1
secθ +
s −2d tanθ
v
2
=
2d
v
1
v
2
_
v
2
2
−v
2
1
+
s −2d
v
1
_
v
2
2
−v
2
1
v
2
=
2d
v
1

v
2
_
v
2
2
−v
2
1

v
2
1
v
2
_
v
2
2
−v
2
1

⎠+
s
v
2
=
2d
v
1

v
2
2
−v
2
1
v
2
_
v
2
2
−v
2
1

⎠+
s
v
2
=
2d
v
1

_
v
2
2
−v
2
1
_
v
2
2

⎠+
s
v
2
=
2d
v
1
_
1−
_
v
1
v
2
_
2
+
s
v
2
=
2d
_
1−k
2
_
1/2
v
1
+
s
v
2
.
(c) Recall that t
1
=
s
v
1
. Wethereforehave
t
2
t
1
=
2d
_
1−k
2
_
1/2
v
1
+
s
v
2
s
v
1
=
2d
_
1−k
2
_
1/2
s
+
v
1
v
2
=
2d
_
1−k
2
_
1/2
s
+k.
Continuewiththeassumptionof thepreviousexercise.
(a) Findthethicknessof thesoil layer, assumingthat v
1
= 0.7v
2
, t
2
/t
1
= 1.3, ands = 400m.
(b) Thetimes t
1
and t
2
aremeasured experimentally. Theequation in Exercise67(c) shows that t
2
/t
1
is alinear
functionof 1/s.Whatmightyouconcludeif experimentswereformedforseveral valuesof s andthepoints(1/s, t
2
/t
1
)
didnot lieonastraight line?
69. InthisexerciseweuseFigure38toproveHeron’sprincipleof Example6withoutcalculus. Bydeﬁnition, C
isthereﬂectionof B acrossthelineMN (sothat BC isperpendicular toMN andBN = CN. Let P betheintersection
of AC andMN. Usegeometrytojustify:
(a) PNB andPNC arecongruent andθ
1
= θ
2
.
(b) ThepathsAPB andAPC haveequal length.
(c) SimilarlyAQB andAQC haveequal length.
(d) ThepathAPC isshorter thanAQC for all Q = P.
Concludethat theshortest pathAQB occursfor Q = P.
A
B
h
1
h
2
P
h
2
Q
C
M N
θ
1
θ
1
θ
2
FIGURE 38
June 9, 2011 LTSV SSM Second Pass
236 C HA P T E R 4 APPLICATIONS OF THE DERIVATIVE
solution
(a) Bydeﬁnition, BC isorthogonal toQM, sotrianglesPNB andPNC arecongruentbyside–angle–side.Therefore
θ
1
= θ
2
(b) BecausePNB andPNC arecongruent, it followsthat PB andPC areof equal length. Thus, pathsAPB and
APC haveequal length.
(c) Thesamereasoningusedinparts (a) and(b) leadus to concludethat QNB andQNC arecongruent andthat
PB andPC areof equal length. Thus, pathsAQB andAQC areof equal length.
(d) Consider triangleAQC. By thetriangleinequality, thelengthof sideAC is less thanor equal to thesumof the
lengthsof thesidesAQandQC. Thus, thepathAPC isshorter thanAQC for all Q = P.
Finally, theshortest pathAQB occursfor Q = P.
A jewelrydesigner planstoincorporateacomponent madeof goldintheshapeof afrustumof aconeof height
1cmandﬁxedlower radiusr (Figure39). Theupper radiusx cantakeonanyvaluebetween0andr. Notethatx = 0
andx = r correspondtoaconeandcylinder, respectively. Asafunctionof x, thesurfacearea(not includingthetop
andbottom) isS(x) = πs(r +x), wheres istheslant height asindicatedintheﬁgure. Whichvalueof x yieldsthe
least expensivedesign[theminimumvalueof S(x) for 0≤ x ≤ r]?
(a) Showthat S(x) = π(r +x)
_
1+(r −x)
2
.
(b) Showthat if r <

2, thenS(x) isanincreasingfunction. Concludethat thecone(x = 0) hasminimal areain
thiscase.
(c) Assumethatr >

2. ShowthatS(x) hastwocritical pointsx
1
< x
2
in(0, r), andthatS(x
1
) isalocal maximum,
andS(x
2
) isalocal minimum.
(d) Concludethat theminimumoccursat x = 0or x
2
.
(e) Findtheminimuminthecasesr = 1.5andr = 2.
(f) Challenge: Let c =
_
(5+3

3)/4 ≈ 1.597. Provethat theminimumoccursat x = 0(cone) if

2 < r < c,
but theminimumoccursat x = x
2
if r > c.
4.7 Newton’s Method
Preliminary Questions
1. Howmanyiterationsof Newton’sMethodarerequiredtocomputearoot if f (x) isalinear function?
solution Remember that Newton’sMethodusesthelinear approximationof afunctiontoestimatethelocationof a
root. If theoriginal functionislinear, thenonlyoneiterationof Newton’sMethodwill berequiredtocomputetheroot.
2. What happensinNewton’sMethodif your initial guesshappenstobeazeroof f ?
solution If x
0
happenstobeazeroof f, then
x
1
= x
0

f (x
0
)
f

(x
0
)
= x
0
−0= x
0
;
inother words, everytermintheNewton’sMethodsequencewill remainx
0
.
3. What happensinNewton’sMethodif your initial guesshappenstobealocal minor maxof f ?
solution Assuming that the function is differentiable, then the derivative is zero at a local maximumor a local
minimum. If Newton’sMethodisstartedwithaninitial guesssuchthat f

(x
0
) = 0, thenNewton’sMethodwill fail in
thesensethat x
1
will not bedeﬁned. That is, thetangent linewill beparallel tothex-axisandwill never intersect it.
4. Isthefollowingareasonabledescriptionof Newton’sMethod: “A root of theequationof thetangent linetof (x) is
usedasanapproximationtoaroot of f (x) itself”? Explain.
solution Yes, thatisareasonabledescription. Theiterationformulafor Newton’sMethodwasderivedbysolvingthe
equationof thetangent linetoy = f (x) at x
0
for itsx-intercept.
Exercises
In this exercise set, all approximations should be carried out using Newton’s Method.
In Exercises 1–6, apply Newton’s Method to f (x) and initial guess x
0
to calculate x
1
, x
2
, x
3
.
1. f (x) = x
2
−6, x
0
= 2
solution Let f (x) = x
2
−6anddeﬁne
x
n+1
= x
n

f (x
n
)
f

(x
n
)
= x
n

x
2
n
−6
2x
n
.
Withx
0
= 2, wecompute
n 1 2 3
x
n
2.5 2.45 2.44948980
f (x) = x
2
−3x +1, x
0
= 3
3. f (x) = x
3
−10, x
0
= 2
solution Let f (x) = x
3
−10anddeﬁne
x
n+1
= x
n

f (x
n
)
f

(x
n
)
= x
n

x
3
n
−10
3x
2
n
.
Withx
0
= 2wecompute
n 1 2 3
x
n
2.16666667 2.15450362 2.15443469
June 9, 2011 LTSV SSM Second Pass
S E C T I ON 4.7 Newton’s Method 237
f (x) = x
3
+x +1, x
0
= −1
5. f (x) = cosx −4x, x
0
= 1
solution Let f (x) = cosx −4x anddeﬁne
x
n+1
= x
n

f (x
n
)
f

(x
n
)
= x
n

cosx
n
−4x
n
−sinx
n
−4
.
Withx
0
= 1wecompute
n 1 2 3
x
n
0.28540361 0.24288009 0.24267469
f (x) = 1−x sinx, x
0
= 7
7. UseFigure6to chooseaninitial guess x
0
to theuniquereal root of x
3
+ 2x + 5 = 0andcomputetheﬁrst three
Newtoniterates.
2 1 −2 −1
x
y
FIGURE 6 Graphof y = x
3
+2x +5.
solution Let f (x) = x
3
+2x +5anddeﬁne
x
n+1
= x
n

f (x
n
)
f

(x
n
)
= x
n

x
3
n
+2x
n
+5
3x
2
n
+2
.
Wetakex
0
= −1.4, basedontheﬁgure, andthencalculate
n 1 2 3
x
n
−1.330964467 −1.328272820 −1.328268856
Approximate a solution of sinx = cos2x in the interval
_
0,
π
2
_
to three decimal places. Then ﬁnd the exact
solutionandcomparewithyour approximation.
9. Approximatethepoint of intersectionof thegraphsy = x
2
+4+
1
x
andy =
2
x
2
tothreedecimal places(Figure7).
1 2
5
10
y
x
y = 2/x
2
y = x
2
+ 4 + 1/x
FIGURE 7 Graphsof x
2
+4+
1
x
and
2
x
2
.
solution Thepoint of intersectionisthesolutionof f (x) = x
2
+4+x
−1
+2x
−2
. Weuseaninitial guessof 0.5.
Newton’sMethod x
0
= 0.5(guess)
x
1
= 0.5−
f (0.5)
f

(0.5)
x
1
≈ 0.560345
x
2
= 0.560345−
f (0.560345)
f

(0.560345)
x
2
≈ 0.573460
x
3
= 0.573460−
f (0.573460)
f

(0.573460)
x
3
≈ 0.573927
x
4
= 0.573927−
f (0.573927)
f

(0.573927)
x
4
≈ 0.573928
Thepointof intersectionhasx-coordinate≈ 0.574tothreedecimal places. Thecorrespondingy-coordinateis
2
0.574
2

6.070.
Theﬁrst positivesolutionof sinx = 0isx = π. UseNewton’sMethodtocalculateπ tofour decimal places.
June 9, 2011 LTSV SSM Second Pass
238 C HA P T E R 4 APPLICATIONS OF THE DERIVATIVE
In Exercises 11–14, approximate to three decimal places using Newton’s Method and compare with the value from a
calculator.
11.

11
solution Let f (x) = x
2
−11, andlet x
0
= 3. Newton’sMethodyields:
n 1 2 3
x
n
3.33333333 3.31666667 3.31662479
A calculator yields3.31662479.
5
1/3
13. 2
7/3
solution Notethat 2
7/3
= 4· 2
1/3
. Let f (x) = x
3
−2, andlet x
0
= 1. Newton’sMethodyields:
n 1 2 3
x
n
1.33333333 1.26388889 1.25993349
Thus, 2
7/3
≈ 4· 1.25993349= 5.03973397. A calculator yields5.0396842.
3
−1/4
15. Approximatethelargest positiveroot of f (x) = x
4
− 6x
2
+ x + 5to within an error of at most 10
−4
. Refer to
Figure5.
solution Figure5 fromthetext suggests thelargest positiveroot of f (x) = x
4
− 6x
2
+ x + 5 is near 2. So let
f (x) = x
4
−6x
2
+x +5andtakex
0
= 2.
n 1 2 3 4
x
n
2.111111111 2.093568458 2.093064768 2.093064358
Thelargest positiveroot of x
4
−6x
2
+x +5isapproximately2.093064358.
In Exercises 16–19, approximate the root speciﬁed to three decimal places using Newton’s Method. Use a plot to
choose an initial guess.
Largest positiveroot of f (x) = x
3
−5x +1.
17. Negativeroot of f (x) = x
5
−20x +10.
solution Let f (x) = x
5
− 20x + 10. Thegraphof f (x) shownbelowsuggests takingx
0
= −2.2. Startingfrom
x
0
= −2.2, theﬁrst threeiteratesof Newton’sMethodare:
n 1 2 3
x
n
−2.22536529 −2.22468998 −2.22468949
Thus, tothreedecimal places, thenegativeroot of f (x) = x
5
−20x +10is−2.225.
−150
−100
−50
−2.5 −2.0 −1.5 −1.0 −0.5
y
x
Positivesolutionof sinθ = 0.8θ.
19. Positivesolutionof 4cosx = x
2
.
solution Fromthe graph below, we see that the positive solution to the equation 4cosx = x
2
is approximately
x = 1.2. Choosingx
0
= 1.2, theﬁrst twoiteratesof Newton’sMethodappliedtof (x) = 4cosx −x
2
are
n 1 2
x
n
1.20154 1.20154
Thus, tothreedecimal places, thepositivesolutionto4cosx = x
2
is1.20154.
June 9, 2011 LTSV SSM Second Pass
S E C T I ON 4.7 Newton’s Method 239
1
0
2
0.5 1.0 1.5 2.0
3
4
5
6
x
y
4 cos x
x
2
Let x
1
, x
2
betheestimatestoaroot obtainedbyapplyingNewton’sMethodwithx
0
= 1tothefunctiongraphed
inFigure8. Estimatethenumerical valuesof x
1
andx
2
, anddrawthetangent linesusedtoobtainthem.
21. Findthesmallest positivevalueof x at whichy = x andy = tanx intersect. Hint: Drawaplot.
solution Hereisaplot of tanx andx onthesameaxes:
4 3 1 2
x
y
−5
5
The ﬁrst intersection with x > 0 lies on the second “branch” of y = tanx, between x =

4
and x =

2
. Let
f (x) = tanx −x. Thegraphsuggestsaninitial guessx
0
=

4
, fromwhichweget thefollowingtable:
n 1 2 3 4
x
n
6.85398 21.921 4480.8 7456.27
Thisisclearlyleadingnowhere, soweneedtotryabetter initial guess. Note: This happens with Newton’s Method—it is
sometimes difﬁcult to choose an initial guess. Wetrythepoint directlybetween

4
and

2
, x
0
=
11π
8
:
n 1 2 3 4 5 6 7
x
n
4.64662 4.60091 4.54662 4.50658 4.49422 4.49341 4.49341
Theﬁrst point wherey = x andy = tanx crossisat approximatelyx = 4.49341, whichisapproximately1.4303π.
In 1535, the mathematicianAntonio Fior challenged his rival Niccolo Tartaglia to solve this problem: A tree
stands12braccia high; it isbrokenintotwopartsat suchapoint that theheight of thepart left standingisthecube
root of thelengthof thepart cut away. What is theheight of thepart left standing? Showthat this is equivalent to
solvingx
3
+x = 12andﬁndtheheighttothreedecimal places. Tartaglia, whohaddiscoveredthesecretof thecubic
equation, wasabletodeterminetheexact answer:
x =
_
3
_

2919+54−
3
_

2919−54
__
3

9
23. Find(totwodecimal places) thecoordinatesof thepoint P inFigure9wherethetangent linetoy = cosx passes
throughtheorigin.
P
y = cos x

1
x
y
FIGURE 9
solution Let (x
r
, cos(x
r
)) bethecoordinates of thepoint P. Theslopeof thetangent lineis −sin(x
r
), so weare
lookingfor atangent line:
y = −sin(x
r
)(x −x
r
) +cos(x
r
)
suchthat y = 0whenx = 0. Thisgivesustheequation:
−sin(x
r
)(−x
r
) +cos(x
r
) = 0.
Let f (x) = cosx + x sinx. Wearelookingfor theﬁrst point x = r wheref (r) = 0. Thesketchgivenindicates that
x
0
= 3π/4wouldbeagoodinitial guess. ThefollowingtablegivessuccessiveNewton’sMethodapproximations:
n 1 2 3 4
x
n
2.931781309 2.803636974 2.798395826 2.798386046
Thepoint P hasapproximatecoordinates(2.7984, −0.941684).
June 9, 2011 LTSV SSM Second Pass
240 C HA P T E R 4 APPLICATIONS OF THE DERIVATIVE
Newton’s Method is often used to determine interest rates in ﬁnancial calculations. In Exercises 24–26, r denotes a yearly
interest rate expressed as a decimal (rather than as a percent).
If P dollarsaredepositedeverymonthinanaccount earninginterest at theyearlyrater, thenthevalueS of the
account after N yearsis
S = P
_
b
12N+1
−b
b −1
_
whereb = 1+
r
12
Youhavedecidedtodeposit P = 100dollarsper month.
(a) DetermineS after 5yearsif r = 0.07(that is, 7%).
(b) Showthat to save\$10,000after 5years, youmust earninterest at arater determinedbys theequationb
61

101b +100= 0. UseNewton’sMethodtosolvefor b. Thenﬁndr. Notethat b = 1isaroot, but youwant theroot
satisfyingb > 1.
25. If youborrowL dollarsfor N yearsatayearlyinterestrater, your monthlypaymentof P dollarsiscalculatedusing
theequation
L = P
_
1−b
−12N
b −1
_
whereb = 1+
r
12
(a) FindP if L = \$5000, N = 3, andr = 0.08(8%).
(b) Youareofferedaloanof L = \$5000tobepaidbackover3yearswithmonthlypaymentsof P = \$200. UseNewton’s
Methodtocomputeb andﬁndtheimpliedinterestrater of thisloan. Hint: Showthat(L/P)b
12N+1
−(1+L/P)b
12N
+
1= 0.
solution
(a) b = (1+0.08/12) = 1.00667
P = L
_
b −1
1−b
−12N
_
= 5000
_
1.00667−1
1−1.00667
−36
_
≈ \$156.69
(b) Startingfrom
L = P
_
1−b
−12N
b −1
_
,
dividebyP, multiplybyb −1, multiplybyb
12N
andcollect liketermstoarriveat
(L/P)b
12N+1
−(1+L/P)b
12N
+1= 0.
SinceL/P = 5000/200= 25, wemust solve
25b
37
−26b
36
+1= 0.
Newton’sMethodgivesb ≈ 1.02121and
r = 12(b −1) = 12(0.02121) ≈ 0.25452
Sotheinterest rateisaround25.45%.
If youdeposit P dollars inaretirement fundevery year for N years withtheintentionof thenwithdrawingQ
dollarsperyearforMyears, youmustearninterestatarater satisfyingP(b
N
−1) = Q(1−b
−M
), whereb = 1+r.
Assumethat\$2000isdepositedeachyear for 30yearsandthegoal istowithdraw\$10,000per year for 25years. Use
Newton’sMethodtocomputeb andthenﬁndr. Notethat b = 1isaroot, but youwant theroot satisfyingb > 1.
27. Thereisnosimpleformulafor thepositionat timet of aplanet P initsorbit (anellipse) aroundthesun. Introduce
theauxiliary circleandangleθ inFigure10(notethat P determinesθ becauseit isthecentral angleof point B onthe
circle). Let a = OA ande = OS/OA (theeccentricityof theorbit).
(a) Showthat sector BSA hasarea(a
2
/2)(θ −e sinθ).
(b) By Kepler’sSecondLaw, theareaof sector BSA isproportional tothetimet elapsedsincetheplanet passedpoint
A, andbecausethecirclehasareaπa
2
, BSA hasarea(πa
2
)(t /T ), whereT istheperiodof theorbit. DeduceKepler’s
Equation:
2πt
T
= θ −e sinθ
(c) Theeccentricity of Mercury’s orbit is approximately e = 0.2. UseNewton’s Method to ﬁnd θ after aquarter of
Mercury’s year has elapsed(t = T/4). Convert θ to degrees. Has Mercury coveredmorethanaquarter of its orbit at
t = T/4?
O
P
A
S
Auxiliary circle
Elliptical orbit
Sun
q
B
FIGURE 10
June 9, 2011 LTSV SSM Second Pass
S E C T I ON 4.7 Newton’s Method 241
solution
(a) Thesector SAB is thesliceOAB withthetriangleOPS removed. OAB is acentral sector witharc θ andradius
OA = a, andthereforehasarea
a
2
θ
2
. OPS isatrianglewithheight a sinθ andbaselengthOS = ea. Hence, theareaof
thesector is
a
2
2
θ −
1
2
ea
2
sinθ =
a
2
2
(θ −e sinθ).
(b) SinceKepler’s secondlawindicates that theareaof thesector is proportional tothetimet sincetheplanet passed
point A, weget
πa
2
(t /T ) = a
2
/2(θ −e sinθ)

t
T
= θ −e sinθ.
(c) If t = T/4, thelast equationin(b) gives:
π
2
= θ −e sinθ = θ −0.2sinθ.
Let f (θ) = θ − 0.2sinθ −
π
2
. Wewill useNewton’sMethodtoﬁndthepoint wheref (θ) = 0. Sinceaquarter of the
year onMercuryhaspassed, agoodﬁrst estimateθ
0
wouldbe
π
2
.
n 1 2 3 4
x
n
1.7708 1.76696 1.76696 1.76696
Fromthepointof viewof theSun, Mercuryhastraversedanangleof approximately1.76696radians = 101.24

. Mercury
hasthereforetraveledmorethanonefourthof thewayaround(fromthepoint of viewof central angle) duringthistime.
Therootsof f (x) =
1
3
x
3
−4x +1tothreedecimal placesare−3.583, 0.251, and3.332(Figure11). Determine
theroottowhichNewton’sMethodconvergesfor theinitial choicesx
0
= 1.85, 1.7, and1.55. Theanswer showsthat
asmall changeinx
0
canhaveasigniﬁcant effect ontheoutcomeof Newton’sMethod.
29. What happenswhenyouapplyNewton’sMethodtoﬁndazeroof f (x) = x
1/3
? Notethat x = 0istheonlyzero.
solution Let f (x) = x
1/3
. Deﬁne
x
n+1
= x
n

f (x
n
)
f

(x
n
)
= x
n

x
1/3
n
1
3
x
−2/3
n
= x
n
−3x
n
= −2x
n
.
Takex
0
= 0.5. Thenthesequenceof iteratesis−1, 2, −4, 8, −16, 32, −64, . . . That is, for any nonzerostartingvalue,
thesequenceof iteratesdivergesspectacularly, sincex
n
= (−2)
n
x
0
. Thuslim
n→∞
|x
n
| = lim
n→∞
2
n
|x
0
| = ∞.
What happens whenyouapply Newton’s Methodto theequationx
3
− 20x = 0withtheunlucky initial guess
x
0
= 2?
Further Insights and Challenges
31. Newton’sMethodcanbeusedtocomputereciprocalswithoutperformingdivision.Letc > 0andsetf (x) = x
−1
−c.
(a) Showthat x −(f (x)/f

(x)) = 2x −cx
2
.
(b) Calculatetheﬁrstthreeiteratesof Newton’sMethodwithc = 10.3andthetwoinitial guessesx
0
= 0.1andx
0
= 0.5.
(c) Explaingraphicallywhyx
0
= 0.5doesnot yieldasequenceconvergingto1/10.3.
solution
(a) Let f (x) =
1
x
−c. Then
x −
f (x)
f

(x)
= x −
1
x
−c
−x
−2
= 2x −cx
2
.
(b) For c = 10.3, wehavef (x) =
1
x
−10.3andthusx
n+1
= 2x
n
−10.3x
2
n
.
• Takex
0
= 0.1.
n 1 2 3
x
n
0.097 0.0970873 0.09708738
• Takex
0
= 0.5.
n 1 2 3
x
n
−1.575 −28.7004375 −8541.66654
(c) Thegraphisdisconnected. If x
0
= 0.5, (x
1
, f (x
1
)) isontheother portionof thegraph, whichwill never converge
toanypoint under Newton’sMethod.
June 9, 2011 LTSV SSM Second Pass
242 C HA P T E R 4 APPLICATIONS OF THE DERIVATIVE
In Exercises 32 and 33, consider a metal rod of length Lfastened at both ends. If you cut the rod and weld on an additional
segment of length m, leaving the ends ﬁxed, the rod will bow up into a circular arc of radius R (unknown), as indicated
in Figure 12.
R
h
q
L
FIGURE 12 Theboldcircular archaslengthL +m.
Let h bethemaximumvertical displacement of therod.
(a) Showthat L = 2Rsinθ andconcludethat
h =
L(1−cosθ)
2sinθ
(b) Showthat L +m = 2Rθ andthenprove
sinθ
θ
=
L
L +m
33. Let L = 3andm = 1. ApplyNewton’sMethodtoEq. (2) toestimateθ, andusethistoestimateh.
solution Welet L = 3andm = 1. Wewant thesolutionof:
sinθ
θ
=
L
L +m
sinθ
θ

L
L +m
= 0
sinθ
θ

3
4
= 0.
Let f (θ) =
sinθ
θ

3
4
.
1.5 0.5 1
x
y
−0.2
−0.2
0.2
0.1
Theﬁgureabovesuggeststhat θ
0
= 1.5wouldbeagoodinitial guess. TheNewton’sMethodapproximationsfor the
solutionfollow:
n 1 2 3 4
θ
n
1.2854388 1.2757223 1.2756981 1.2756981
Theanglewhere
sinθ
θ
=
L
L+m
isapproximately1.2757. Hence
h = L
1−cosθ
2sinθ
≈ 1.11181.
QuadraticConvergencetoSquareRoots Let f (x) = x
2
−c andlet e
n
= x
n

c betheerror inx
n
.
(a) Showthat x
n+1
=
1
2
(x
n
+c/x
n
) ande
n+1
= e
2
n
/2x
n
.
(b) Showthat if x
0
>

c, thenx
n
>

c for all n. Explaingraphically.
(c) Showthat if x
0
>

c, thene
n+1
≤ e
2
n
/(2

c).
4.8 Antiderivatives
Preliminary Questions
1. Findanantiderivativeof thefunctionf (x) = 0.
solution Since the derivative of any constant is zero, any constant function is an antiderivative for the function
f (x) = 0.
2. Isthereadifferencebetweenﬁndingthegeneral antiderivativeof afunctionf (x) andevaluating
_
f (x) dx?
solution Nodifference. Theindeﬁniteintegral isthesymbol for denotingthegeneral antiderivative.
3. J acqueswastoldthat f (x) andg(x) havethesamederivative, andhewonderswhether f (x) = g(x). DoesJ acques
havesufﬁcient informationtoanswer hisquestion?
solution No. Knowingthat thetwofunctionshavethesamederivativeisonly goodenoughtotell J acquesthat the
functionsmaydiffer byatmostanadditiveconstant. Todeterminewhether thefunctionsareequal for all x, J acquesneeds
toknowthevalueof eachfunctionfor asinglevalueof x. If thetwofunctionsproducethesameoutput valuefor asingle
input value, theymust takethesamevaluefor all input values.
4. Supposethat F

(x) = f (x) andG

(x) = g(x). Whichof thefollowingstatementsaretrue? Explain.
(a) If f = g, thenF = G.
(b) If F andG differ byaconstant, thenf = g.
(c) If f andg differ byaconstant, thenF = G.
June 9, 2011 LTSV SSM Second Pass
S E C T I ON 4.8 Antiderivatives 243
solution
(a) False. Evenif f (x) = g(x), theantiderivativesF andG maydiffer byanadditiveconstant.
(b) True. Thisfollowsfromthefact that thederivativeof anyconstant is0.
(c) False. If thefunctionsf andg aredifferent, thentheantiderivativesFandGdifferbyalinearfunction: F(x) −G(x) =
ax +b for someconstantsa andb.
5. Isy = x asolutionof thefollowingInitial ValueProblem?
dy
dx
= 1, y(0) = 1
solution Although
d
dx
x = 1, thefunctionf (x) = x takesthevalue0whenx = 0, soy = x isnot asolutionof the
indicatedinitial valueproblem.
Exercises
In Exercises 1–8, ﬁnd the general antiderivative of f (x) and check your answer by differentiating.
1. f (x) = 18x
2
solution
_
18x
2
dx = 18
_
x
2
dx = 18·
1
3
x
3
+C = 6x
3
+C.
Asacheck, wehave
d
dx
(6x
3
+C) = 18x
2
asneeded.
f (x) = x
−3/5
3. f (x) = 2x
4
−24x
2
+12x
−1
solution
_
(2x
4
−24x
2
+12x
−1
) dx = 2
_
x
4
dx −24
_
x
2
dx +12
_
1
x
dx
= 2·
1
5
x
5
−24·
1
3
x
3
+12ln|x| +C
=
2
5
x
5
−8x
3
+12ln|x| +C.
Asacheck, wehave
d
dx
_
2
5
x
5
−8x
3
+12ln|x| +C
_
= 2x
4
−24x
2
+12x
−1
asneeded.
f (x) = 9x +15x
−2
5. f (x) = 2cosx −9sinx
solution
_
(2cosx −9sinx) dx = 2
_
cosx dx −9
_
sinx dx
= 2sinx −9(−cosx) +C = 2sinx +9cosx +C
Asacheck, wehave
d
dx
(2sinx +9cosx +C) = 2cosx +9(−sinx) = 2cosx −9sinx
asneeded.
f (x) = 4x
7
−3cosx
7. f (x) = sin2x +12cos3x
solution
_
(sin2x +12cos3x) dx =
_
sin2x dx +12
_
cos3x dx = −
1
2
cos2x +
12
3
sin3x +C
= 4sin3x −
1
2
cos2x +C
Asacheck, wehave
d
dx
_
4sin3x −
1
2
cos2x +C
_
= sin2x +12cos3x
June 9, 2011 LTSV SSM Second Pass
244 C HA P T E R 4 APPLICATIONS OF THE DERIVATIVE
f (x) = sin(4−9x)
9. Matchfunctions(a)–(d) withtheir antiderivatives(i)–(iv).
(a) f (x) = sinx (i) F(x) = cos(1−x)
(b) f (x) = x sin(x
2
) (ii) F(x) = −cosx
(c) f (x) = sin(1−x) (iii) F(x) = −
1
2
cos(x
2
)
(d) f (x) = x sinx (iv) F(x) = sinx −x cosx
solution
(a) Anantiderivativeof sinx is−cosx, whichis(ii). Asacheck, wehave
d
dx
(−cosx) = −(−sinx) = sinx.
(b) An antiderivative of x sin(x
2
) is −
1
2
cos(x
2
), which is (iii). This is because, by the Chain Rule, we have
d
dx
_

1
2
cos(x
2
)
_
= −
1
2
_
−sin(x
2
)
_
· 2x = x sin(x
2
).
(c) Anantiderivativeof sin(1−x) iscos(1−x) or (i). Asacheck, wehave
d
dx
cos(1− x) = −sin(1− x) · (−1) =
sin(1−x).
(d) Anantiderivativeof x sinx issinx −x cosx, whichis(iv). Thisisbecause
d
dx
(sinx −x cosx) = cosx −(x (−sinx) +cosx · 1) = x sinx
In Exercises 10–39, evaluate the indeﬁnite integral.
_
(9x +2) dx
11.
_
(4−18x) dx
solution
_
(4−18x) dx = 4x −9x
2
+C.
_
x
−3
dx
13.
_
t
−6/11
dt
solution
_
t
−6/11
dt =
t
5/11
5/11
+C =
11
5
t
5/11
+C.
_
(5t
3
−t
−3
) dt
15.
_
(18t
5
−10t
4
−28t ) dt
solution
_
(18t
5
−10t
4
−28t ) dt = 3t
6
−2t
5
−14t
2
+C.
_
14s
9/5
ds
17.
_
(z
−4/5
−z
2/3
+z
5/4
) dz
solution
_
((z
−4/5
−z
2/3
+z
5/4
) dz =
z
1/5
1/5

z
5/3
5/3
+
z
9/4
9/4
+C = 5z
1/5

3
5
z
5/3
+
4
9
z
9/4
+C.
_
3
2
dx
19.
_
1
3

x
dx
solution
_
1
3

x
dx =
_
x
−1/3
dx =
x
2/3
2/3
+C =
3
2
x
2/3
+C.
_
dx
x
4/3
21.
_
36dt
t
3
solution
_
36
t
3
dt =
_
36t
−3
dt = 36
t
−2
−2
+C = −
18
t
2
+C.
_
x(x
2
−4) dx
23.
_
(t
1/2
+1)(t +1) dt
solution
_
(t
1/2
+1)(t +1) dt =
_
(t
3/2
+t +t
1/2
+1) dt
=
t
5/2
5/2
+
1
2
t
2
+
t
3/2
3/2
+t +C
=
2
5
t
5/2
+
1
2
t
2
+
2
3
t
3/2
+t +C
June 9, 2011 LTSV SSM Second Pass
S E C T I ON 4.8 Antiderivatives 245
_
12−z

z
dz
25.
_
x
3
+3x
2
−4
x
2
dx
solution
_
x
3
+3x
2
−4
x
2
dx =
_
(x +3−4x
−2
) dx
=
1
2
x
2
+3x +4x
−1
+C
_ _
1
3
sinx −
1
4
cosx
_
dx
27.
_
12secx tanx dx
solution
_
12secx tanx dx = 12secx +C.
_
(θ +sec
2
θ) dθ
29.
_
(csct cott ) dt
solution
_
(csct cott ) dt = −csct +C.
_
sin(7x −5) dx
31.
_
sec
2
(7−3θ) dθ
solution
_
sec
2
(7−3θ) dθ = −
1
3
tan(7−3θ) +C.
_
(θ −cos(1−θ)) dθ
33.
_
25sec
2
(3z +1) dz
solution
_
25sec
2
(3z +1) dz =
25
3
tan(3z +1) +C.
_
(12cos4x +9sin3x) dx
35.
_
sec12t tan12t dt
solution
_
sec12t tan12t dt =
1
12
sec12t +C
_
5tan(4θ +3) sec(4θ +3) dθ
37.
_
sec4x(3sec4x −5tan4x) dx
solution
_
sec4x(3sec4x −5tan4x) dx = 3
_
sec
2
4x dx −5
_
sec4x tan4x dx =
3
4
tan4x −
5
4
sec4x +C
_
sec(x +5) tan(x +5) dx
39.
_ _
cos(3θ) −
1
2
sec
2
_
θ
4
__

solution
_ _
cos(3θ) −
1
2
sec
2
_
θ
4
__
dθ =
1
3
sin(3θ) −2tan
_
θ
4
_
+C.
InFigure2, isgraph(A) or graph(B) thegraphof anantiderivativeof f (x)?
41. InFigure3, whichof graphs(A), (B), and(C) isnot thegraphof anantiderivativeof f (x)? Explain.
f (x)
(C) (B) (A)
x
x
y
x
y
x
y
y
FIGURE 3
solution Let F(x) bean antiderivativeof f (x). Noticethat f (x) = F

(x) changes sign from− to + to − to +.
Hence, F(x) must transitionfromdecreasingtoincreasingtodecreasingtoincreasing.
• Bothgraph(A) andgraph(C) meet thecriteriadiscussedaboveandonlydiffer byanadditiveconstant. Thuseither
couldbeanantiderivativeof f (x).
• Graph(B) doesnothavethesamelocal extremaasindicatedbyf (x) andthereforeisnot anantiderivativeof f (x).
June 9, 2011 LTSV SSM Second Pass
246 C HA P T E R 4 APPLICATIONS OF THE DERIVATIVE
Showthat F(x) =
1
3
(x +13)
3
isanantiderivativeof f (x) = (x +13)
2
.
In Exercises 43–46, verify by differentiation.
43.
_
(x +13)
6
dx =
1
7
(x +13)
7
+C
solution
d
dx
_
1
7
(x +13)
7
+C
_
= (x +13)
6
asrequired.
_
(x +13)
−5
dx = −
1
4
(x +13)
−4
+C
45.
_
(4x +13)
2
dx =
1
12
(4x +13)
3
+C
solution
d
dx
_
1
12
(4x +13)
3
+C
_
=
1
4
(4x +13)
2
(4) = (4x +13)
2
asrequired.
_
(ax +b)
n
dx =
1
a(n +1)
(ax +b)
n+1
+C (for n = −1)
In Exercises 47–62, solve the initial value problem.
47.
dy
dx
= x
3
, y(0) = 4
solution Since
dy
dx
= x
3
, wehave
y =
_
x
3
dx =
1
4
x
4
+C.
Thus,
4= y(0) =
1
4
0
4
+C = C,
sothat C = 4. Therefore, y =
1
4
x
4
+4.
dy
dt
= 3−2t , y(0) = −5
49.
dy
dt
= 2t +9t
2
, y(1) = 2
solution Since
dy
dt
= 2t +9t
2
, wehave
y =
_
(2t +9t
2
) dt = t
2
+3t
3
+C.
Thus,
2= y(1) = 1
2
+3(1)
3
+C,
sothat C = −2. Thereforey = t
2
+3t
3
−2.
dy
dx
= 8x
3
+3x
2
, y(2) = 0
51.
dy
dt
=

t , y(1) = 1
solution Since
dy
dt
=

t = t
1/2
, wehave
y =
_
t
1/2
dt =
2
3
t
3/2
+C.
Thus
1= y(1) =
2
3
+C,
sothat C =
1
3
. Therefore, y =
2
3
t
3/2
+
1
3
.
dz
dt
= t
−3/2
, z(4) = −1
53.
dy
dx
= (3x +2)
3
, y(0) = 1
solution Since
dy
dx
= (3x +2)
3
, wehave
y =
_
(3x +2)
3
dx =
1
4
·
1
3
(3x +2)
4
+C =
1
12
(3x +2)
4
+C.
Thus,
1= y(0) =
1
12
(2)
4
+C,
sothat C = 1−
4
3
= −
1
3
. Therefore, y =
1
12
(3x +2)
4

1
3
.
dy
dt
= (4t +3)
−2
, y(1) = 0
June 9, 2011 LTSV SSM Second Pass
S E C T I ON 4.8 Antiderivatives 247
55.
dy
dx
= sinx, y
_
π
2
_
= 1
solution Since
dy
dx
= sinx, wehave
y =
_
sinx dx = −cosx +C.
Thus
1= y
_
π
2
_
= 0+C,
sothat C = 1. Therefore, y = 1−cosx.
dy
dz
= sin2z, y
_
π
4
_
= 4
57.
dy
dx
= cos5x, y(π) = 3
solution Since
dy
dx
= cos5x, wehave
y =
_
cos5x dx =
1
5
sin5x +C.
Thus3= y(π) = 0+C, sothat C = 3. Therefore, y = 3+
1
5
sin5x.
dy
dx
= sec
2
3x, y
_
π
4
_
= 2
59.
dy

= 6sec3θ tan3θ, y
_
π
12
_
= −4
solution Wehave
y =
_
6sec3θ tan3θ dθ = 2sec3θ +C
Thus,
−4= y
_
π
12
_
= 2sec3·
π
12
+C = 2sec
π
4
+C = 2

2+C
sothat C = −4−2

2. Therefore, y = 2sec3θ −4−2

2.
dy
dt
= 4t −sin2t , y(0) = 2
61.
dy

= cos
_
3π −
1
2
θ
_
, y(3π) = 8.
solution Wehave
y =
_
cos
_
3π −
1
2
θ
_
dθ = −2sin
_
3π −
1
2
θ
_
+C
Thus,
8= y(3π) = −2sin
_
3π −

2
_
+C = −2sin
_

2
_
+C = 2+C
sothat C = 6. Therefore, y = −2sin
_
3π −
1
2
θ
_
+6
dy
dx
=
1
x
2
−csc
2
x, y
_
π
2
_
= 0
In Exercises 63–69, ﬁrst ﬁnd f

and then ﬁnd f .
63. f

(x) = 12x, f

(0) = 1, f (0) = 2
solution Let f

(x) = 12x. Thenf

(x) = 6x
2
+ C. Givenf

(0) = 1, it follows that 1 = 6(0)
2
+ C andC = 1.
Thus, f

(x) = 6x
2
+ 1. Next, f (x) = 2x
3
+ x + C. Givenf (0) = 2, it followsthat 2 = 2(0)
3
+ 0+ C andC = 2.
Finally, f (x) = 2x
3
+x +2.
f

(x) = x
3
−2x, f

(1) = 0, f (1) = 2
65. f

(x) = x
3
−2x +1, f

(0) = 1, f (0) = 0
solution Let g(x) = f

(x). Thestatement gives us g

(x) = x
3
− 2x + 1, g(0) = 1. Fromthis, weget g(x) =
1
4
x
4
− x
2
+ x + C. g(0) = 1givesus1= C, sof

(x) = g(x) =
1
4
x
4
− x
2
+ x + 1. f

(x) =
1
4
x
4
− x
2
+ x + 1, so
f (x) =
1
20
x
5

1
3
x
3
+
1
2
x
2
+x +C. f (0) = 0givesC = 0, so
f (x) =
1
20
x
5

1
3
x
3
+
1
2
x
2
+x.
f

(x) = x
3
−2x +1, f

(1) = 0, f (1) = 4
67. f

(t ) = t
−3/2
, f

(4) = 1, f (4) = 4
solution Letg(t ) = f

(t ).Theproblemstatementisg

(t ) = t
−3/2
, g(4) = 1.Fromg

(t ) wegetg(t ) =
1
−1/2
t
−1/2
+
C = −2t
−1/2
+ C. Fromg(4) = 1weget −1+ C = 1so that C = 2. Hencef

(t ) = g(t ) = −2t
−1/2
+ 2. From
f

(t ) weget f (t ) = −2
1
1/2
t
1/2
+2t +C = −4t
1/2
+2t +C. Fromf (4) = 4weget −8+8+C = 4, sothat C = 4.
Hence, f (t ) = −4t
1/2
+2t +4.
June 9, 2011 LTSV SSM Second Pass
248 C HA P T E R 4 APPLICATIONS OF THE DERIVATIVE
f

(θ) = cosθ, f

_
π
2
_
= 1, f
_
π
2
_
= 6
69. f

(t ) = t −cost , f

(0) = 2, f (0) = −2
solution Let g(t ) = f

(t ). Theproblemstatement gives
g

(t ) = t −cost, g(0) = 2.
Fromg

(t ), wegetg(t ) =
1
2
t
2
−sint +C. Fromg(0) = 2, wegetC = 2. Hencef

(t ) = g(t ) =
1
2
t
2
−sint +2. From
f

(t ), weget f (t ) =
1
2
(
1
3
t
3
) +cost +2t +C. Fromf (0) = −2, weget 1+C = −2, henceC = −3, and
f (t ) =
1
6
t
3
+cost +2t −3.
ShowthatF(x) = tan
2
x andG(x) = sec
2
x havethesamederivative. Whatcanyouconcludeabouttherelation
betweenF andG?Verifythisconclusiondirectly.
71. A particlelocated at theorigin at t = 1 s moves along thex-axis with velocity v(t ) = (6t
2
− t ) m/s. Statethe
differential equationwithinitial conditionsatisﬁedbythepositions(t ) of theparticle, andﬁnds(t ).
solution Thedifferential equationsatisﬁedbys(t ) is
ds
dt
= v(t ) = 6t
2
−t,
andtheassociatedinitial conditioniss(1) = 0. Fromthedifferential equation, weﬁnd
s(t ) =
_
(6t
2
−t ) dt = 2t
3

1
2
t
2
+C.
Usingtheinitial condition, it followsthat
0= s(1) = 2−
1
2
+C so C = −
3
2
.
Finally,
s(t ) = 2t
3

1
2
t
2

3
2
.
A particlemovesalongthex-axiswithvelocityv(t ) = (6t
2
− t ) m/s. Findtheparticle’spositions(t ) assuming
that s(2) = 4.
73. Amassoscillatesattheendof aspring. Lets(t ) bethedisplacementof themassfromtheequilibriumpositionattime
t . Assumingthat themassislocatedat theoriginat t = 0andhasvelocity v(t ) = sin(πt /2) m/s, statethedifferential
equationwithinitial conditionsatisﬁedbys(t ), andﬁnds(t ).
solution Thedifferential equationsatisﬁedbys(t ) is
ds
dt
= v(t ) = sin(πt /2),
andtheassociatedinitial conditioniss(0) = 0. Fromthedifferential equation, weﬁnd
s(t ) =
_
sin(πt /2) dt = −
2
π
cos(πt /2) +C.
Usingtheinitial condition, it followsthat
0= s(0) = −
2
π
+C so C =
2
π
.
Finally,
s(t ) =
2
π
(1−cos(πt /2)).
Beginningat t = 0withinitial velocity 4m/s, aparticlemovesinastraight linewithaccelerationa(t ) = 3t
1/2
m/s
2
. Findthedistancetraveledafter 25seconds.
75. A car traveling25m/sbeginstodecelerateat aconstant rateof 4m/s
2
. After howmanysecondsdoesthecar come
toastopandhowfar will thecar havetraveledbeforestopping?
solution Sincetheaccelerationof thecar isaconstant −4m/s
2
, v isgivenbythedifferential equation:
dv
dt
= −4, v(0) = 25.
From
dv
dt
, weget v(t ) =
_
−4dt = −4t + C. Sincev(0)25, C = 25. Fromthis, v(t ) = −4t + 25
m
s
. Toﬁndthetime
until thecar stops, wemust solvev(t ) = 0:
−4t +25= 0
4t = 25
t = 25/4= 6.25s.
June 9, 2011 LTSV SSM Second Pass
S E C T I ON 4.8 Antiderivatives 249
Nowwehaveadifferential equationfor s(t ). Sincewewant toknowhowfar thecar hastraveledfromthebeginningof
itsdecelerationat timet = 0, wehaves(0) = 0bydeﬁnition, so:
ds
dt
= v(t ) = −4t +25, s(0) = 0.
Fromthis, s(t ) =
_
(−4t +25) dt = −2t
2
+25t +C. Sinces(0) = 0, wehaveC = 0, and
s(t ) = −2t
2
+25t.
At stoppingtimet = 0.25s, thecar hastraveled
s(6.25) = −2(6.25)
2
+25(6.25) = 78.125m.
At timet = 1s, aparticleistravelingat 72m/sandbeginstodecelerateat theratea(t ) = −t
−1/2
until it stops.
Howfar doestheparticletravel beforestopping?
77. A 900-kg rocket is released froma space station. As it burns fuel, the rocket’s mass decreases and its velocity
increases. Let v(m) bethevelocity (inmetersper second) asafunctionof massm. Findthevelocity whenm = 729if
dv/dm = −50m
−1/2
. Assumethat v(900) = 0.
solution Since
dv
dm
= −50m
−1/2
, we have v(m) =
_
−50m
−1/2
dm = −100m
1/2
+ C. Thus 0 = v(900) =
−100

900+C = −3000+C, andC = 3000. Therefore, v(m) = 3000−100

m. Accordingly,
v(729) = 3000−100

729= 3000−100(27) = 300meters/sec.
Aswater ﬂowsthroughatubeof radiusR = 10cm, thevelocityv of anindividual water particledependsonlyon
itsdistancer fromthecenterof thetube.Theparticlesatthewallsof thetubehavezerovelocityanddv/dr = −0.06r.
Determinev(r).
79. Verifythelinearitypropertiesof theindeﬁniteintegral statedinTheorem4.
solution ToverifytheSumRule, let F(x) andG(x) beanyantiderivativesof f (x) andg(x), respectively. Because
d
dx
(F(x) +G(x)) =
d
dx
F(x) +
d
dx
G(x) = f (x) +g(x),
it followsthat F(x) +G(x) isanantiderivativeof f (x) +g(x); i.e.,
_
(f (x) +g(x)) dx =
_
f (x) dx +
_
g(x) dx.
ToverifytheMultiplesRule, againlet F(x) beanyantiderivativeof f (x) andlet c beaconstant. Because
d
dx
(cF(x)) = c
d
dx
F(x) = cf (x),
it followsthat cF(x) isandantiderivativeof cf (x); i.e.,
_
(cf (x)) dx = c
_
f (x) dx.
Further Insights and Challenges
Findconstantsc
1
andc
2
suchthat F(x) = c
1
x sinx +c
2
cosx isanantiderivativeof f (x) = x cosx.
81. Findconstantsc
1
andc
2
suchthat F(x) = c
1
x cosx +c
2
sinx isanantiderivativeof f (x) = x sinx.
solution LetF(x) = c
1
x cosx +c
2
sinx. If F(x) istobeanantiderivativeof f (x), thenwemusthavef (x) = F

(x).
Therefore
x sinx = c
1
cosx −c
1
x sinx +c
2
cosx = −c
1
x sinx +(c
1
+c
2
) cosx
Equatingcoefﬁcientsof x sinx yieldsc
1
= −1; equatingcoefﬁcientsof cosx thengivesc
2
= 1.ThusF(x) = −x cosx +
sinx.
SupposethatF

(x) = f (x) andG

(x) = g(x). IsittruethatF(x)G(x) isanantiderivativeof f (x)g(x)?Conﬁrm
or provideacounterexample.
83. Supposethat F

(x) = f (x).
(a) Showthat
1
2
F(2x) isanantiderivativeof f (2x).
(b) Findthegeneral antiderivativeof f (kx) for k = 0.
solution Let F

(x) = f (x).
(a) BytheChainRule, wehave
d
dx
_
1
2
F(2x)
_
=
1
2
F

(2x) · 2= F

(2x) = f (2x).
Thus
1
2
F(2x) isanantiderivativeof f (2x).
June 9, 2011 LTSV SSM Second Pass
250 C HA P T E R 4 APPLICATIONS OF THE DERIVATIVE
(b) For nonzeroconstant k, theChainRulesgives
d
dx
_
1
k
F (kx)
_
=
1
k
F

(kx) · k = F

(kx) = f (kx).
Thus
1
k
F(kx) is an antiderivativeof f (kx). Hencethegeneral antiderivativeof f (kx) is
1
k
F(kx) + C, whereC is a
constant.
Findanantiderivativefor f (x) = |x|.
85. UsingTheorem1, provethat F

(x) = f (x) wheref (x) isapolynomial of degreen −1, thenF(x) isapolynomial
of degreen. Thenprovethat if g(x) isanyfunctionsuchthatg
(n)
(x) = 0, theng(x) isapolynomial of degreeat mostn.
solution SupposeF

(x) = f (x) wheref (x) isapolynomial of degreen −1. Now, weknowthat thederivativeof a
polynomial of degreen isapolynomial of degreen −1, andhenceanantiderivativeof apolynomial of degreen −1isa
polynomial of degreen. Thus, byTheorem1, F(x) candiffer fromapolynomial of degreen byat most aconstant term,
whichisstill apolynomial of degreen. Now, supposethat g(x) isanyfunctionsuchthat g
(n+1)
(x) = 0. Weknowthat
then +1-st derivativeof anypolynomial of degreeat most n iszero, sobyrepeatedapplicationof Theorem1, g(x) can
differ fromapolynomial of degreeatmostn byatmostaconstantterm. Hence, g(x) isapolynomial of degreeatmostn.
ThePowerRuleforantiderivativesdoesnotapplytof (x) = x
−1
. Whichof thegraphsinFigure4couldplausibly
represent anantiderivativeof f (x) = x
−1
?
CHAPTER REVIEW EXERCISES
In Exercises 1–6, estimate using the Linear Approximation or linearization, and use a calculator to estimate the error.
1. 8.1
1/3
−2
solution Letf (x) = x
1/3
, a = 8andx = 0.1. Thenf

(x) =
1
3
x
−2/3
, f

(a) =
1
12
and, bytheLinearApproxima-
tion,
f = 8.1
1/3
−2≈ f

(a)x =
1
12
(0.1) = 0.00833333.
Usingacalculator, 8.1
1/3
−2= 0.00829885. Theerror intheLinearApproximationistherefore
|0.00829885−0.00833333| = 3.445×10
−5
.
1

4.1

1
2
3. 625
1/4
−624
1/4
solution Let f (x) = x
1/4
, a = 625 and x = −1. Then f

(x) =
1
4
x
−3/4
, f

(a) =
1
500
and, by the Linear
Approximation,
f = 624
1/4
−625
1/4
≈ f

(a)x =
1
500
(−1) = −0.002.
Thus625
1/4
−624
1/4
≈ 0.002. Usingacalculator,
625
1/4
−624
1/4
= 0.00200120.
Theerror intheLinearApproximationistherefore
|0.00200120−(0.002)| = 1.201×10
−6
.

101
5.
1
1.02
solution Let f (x) = x
−1
anda = 1. Thenf (a) = 1, f

(x) = −x
−2
andf

(a) = −1. Thelinearizationof f (x) at
a = 1istherefore
L(x) = f (a) +f

(a)(x −a) = 1−(x −1) = 2−x,
and
1
1.02
≈ L(1.02) = 0.98. Usingacalculator,
1
1.02
= 0.980392, sotheerror intheLinearApproximationis
|0.980392−0.98| = 3.922×10
−4
.
5

33
In Exercises 7–12, ﬁnd the linearization at the point indicated.
7. y =

x, a = 25
solution Let y =

x anda = 25. Theny(a) = 5, y

=
1
2
x
−1/2
andy

(a) =
1
10
. Thelinearizationof y at a = 25is
therefore
L(x) = y(a) +y

(a)(x −25) = 5+
1
10
(x −25).
June 9, 2011 LTSV SSM Second Pass
Chapter Review Exercises 251
v(t ) = 32t −4t
2
, a = 2
9. A(r) =
4
3
πr
3
, a = 3
solution Let A(r) =
4
3
πr
3
anda = 3. ThenA(a) = 36π, A

(r) = 4πr
2
andA

(a) = 36π. Thelinearizationof
A(r) at a = 3istherefore
L(r) = A(a) +A

(a)(r −a) = 36π +36π(r −3) = 36π(r −2).
V(h) = 4h(2−h)(4−2h), a = 1
11. P(θ) = sin(3θ +π), a =
π
3
solution WehaveP(a) = sin(2π) = 0, P

(θ) = 3cos(3θ + π), andP

(a) = 3cos2π = 3. Thelinearizationof
P(θ) at a =
π
3
istherefore
L(h) = P

(a)(h −a) +P(a) = 3
_
h −
π
3
_
= 3h −π
R(t ) = tan
_
π
_
t −
1
2
__
, a =
1
4
In Exercises 13–18, use the Linear Approximation.
13. Thepositionof anobjectinlinear motionattimet iss(t ) = 0.4t
2
+(t +1)
−1
. Estimatethedistancetraveledover
thetimeinterval [4, 4.2].
solution Lets(t ) = 0.4t
2
+(t +1)
−1
, a = 4andt = 0.2. Thens

(t ) = 0.8t −(t +1)
−2
ands

(a) = 3.16. Using
theLinearApproximation, thedistancetraveledover thetimeinterval [4, 4.2] isapproximately
s = s(4.2) −s(4) ≈ s

(a)t = 3.16(0.2) = 0.632.
A bondthat pays\$10,000in6yearsisofferedfor saleat apriceP. ThepercentageyieldY of thebondis
Y = 100
_
_
10,000
P
_
1/6
−1
_
Verifythat if P = \$7500, thenY = 4.91%. Estimatethedropinyieldif thepricerisesto\$7700.
15. WhenabuspassfromAlbuquerquetoLosAlamosispricedatp dollars, abuscompanytakesinamonthlyrevenue
of R(p) = 1.5p −0.01p
2
(inthousandsof dollars).
(a) EstimateR if thepricerisesfrom\$50to\$53.
(b) If p = 80, howwill revenuebeaffectedbyasmall increaseinprice? ExplainusingtheLinearApproximation.
solution
(a) If thepriceisraisedfrom\$50to\$53, thenp = 3and
R ≈ R

(50)p = (1.5−0.02(50))(3) = 1.5
Wethereforeestimateanincreaseof \$1500inrevenue.
(b) BecauseR

(80) = 1.5− 0.02(80) = −0.1, theLinear Approximationgives R ≈ −0.1p. A small increasein
pricewouldthusresult inadecreaseinrevenue.
A storesells80MP4playersper weekwhentheplayersarepricedat P = \$75. Estimatethenumber N soldif P
israisedto\$80, assumingthat dN/dP = −4. EstimateN if thepriceisloweredto\$69.
17. Thecircumferenceof asphereismeasuredatC = 100cm. Estimatethemaximumpercentageerror inV if theerror
inC isat most 3cm.
solution Thevolumeof asphereisV =
4
3
πr
3
andthecircumferenceisC = 2πr, wherer istheradiusof thesphere.
Thus, r =
1

C and
V =
4
3
π
_
C

_
3
=
1

2
C
3
.
UsingtheLinearApproximation,
V ≈
dV
dC
C =
1

2
C
2
C,
so
V
V

1

2
C
2
C
1

2
C
3
= 3
C
C
.
WithC = 100cmandC at most 3cm, weestimatethat themaximumpercentageerror inV is3
3
100
= 0.09, or 9%.
Showthat
_
a
2
+b ≈ a +
b
2a
if b issmall. Usethistoestimate

26andﬁndtheerror usingacalculator.
19. VerifytheMVT for f (x) = x
−1/3
on[1, 8].
solution Wehavef (1) = 1
−1/3
= 1andf (8) = 8
−1/3
=
1
2
, sotheMVT saysthat thereisc ∈ [1, 8] suchthat
f

(c) =
f (8) −f (1)
8−1
= −
1
14
June 9, 2011 LTSV SSM Second Pass
252 C HA P T E R 4 APPLICATIONS OF THE DERIVATIVE
Now, f

(x) = −
1
3
x
−4/3
. Solvingfor f

(x) = −
1
14
gives

1
3
x
−4/3
= −
1
14
⇒ x
−4/3
=
3
14
⇒ x
4/3
=
14
3
⇒ x =
_
14
3
_
3/4
≈ 3.175∈ [1, 8]
Showthat f (x) = 2x
3
+2x +sinx +1haspreciselyonereal root.
21. VerifytheMVT for f (x) = x +
1
x
on[2, 5].
solution Ontheinterval [2, 5], f (x) iscontinuousanddifferentiable, sotheMVT applies. Now, f

(x) = 1−
1
x
2
, so
1−
1
c
2
= f

(c) =
f (b) −f (a)
b −a
=
26
5

5
2
5−2
=
9
10
,
or
c
2
= 10 ⇒ c =

10∈ [2, 5]
Supposethat f (1) = 5andf

(x) ≥ 2for x ≥ 1. UsetheMVT toshowthat f (8) ≥ 19.
23. UsetheMVT toprovethat if f

(x) ≤ 2for x > 0andf (0) = 4, thenf (x) ≤ 2x +4for all x ≥ 0.
solution Letx > 0.Becausef iscontinuouson[0, x] anddifferentiableon(0, x),theMeanValueTheoremguarantees
thereexistsac ∈ (0, x) suchthat
f

(c) =
f (x) −f (0)
x −0
or f (x) = f (0) +xf

(c).
Now, wearegiventhat f (0) = 4andthat f

(x) ≤ 2for x > 0. Therefore, for all x ≥ 0,
f (x) ≤ 4+x(2) = 2x +4.
A functionf (x) has derivativef

(x) =
1
x
4
+1
. Whereontheinterval [1, 4] does f (x) takeonits maximum
value?
In Exercises 25–30, ﬁnd the critical points and determine whether they are minima, maxima, or neither.
25. f (x) = x
3
−4x
2
+4x
solution Let f (x) = x
3
−4x
2
+4x. Thenf

(x) = 3x
2
−8x +4= (3x −2)(x −2), sothat x =
2
3
andx = 2are
critical points. Next, f

(x) = 6x − 8, sof

(
2
3
) = −4 < 0andf

(2) = 4 > 0. Therefore, by theSecondDerivative
Test, f (
2
3
) isalocal maximumwhilef (2) isalocal minimum.
s(t ) = t
4
−8t
2
27. f (x) = x
2
(x +2)
3
solution Let f (x) = x
2
(x +2)
3
. Then
f

(x) = 3x
2
(x +2)
2
+2x(x +2)
3
= x(x +2)
2
(3x +2x +4) = x(x +2)
2
(5x +4),
sothat x = 0, x = −2andx = −
4
5
arecritical points. Thesignof theﬁrst derivativeontheintervalssurroundingthe
critical pointsisindicatedinthetablebelow. Basedonthisinformation, f (−2) isneither alocal maximumnor alocal
minimum, f (−
4
5
) isalocal maximumandf (0) isalocal minimum.
Interval (−∞, −2) (−2, −
4
5
) (−
4
5
, 0) (0, ∞)
Signof f

+ + − +
f (x) = x
2/3
(1−x)
29. g(θ) = sin
2
θ +θ
solution Let g(θ) = sin
2
θ +θ. Then
g

(θ) = 2sinθ cosθ +1= 2sin2θ +1,
sothecritical pointsare
θ =

4
+nπ
for all integers n. Becauseg

(θ) ≥ 0 for all θ, it follows that g
_

4
+nπ
_
is neither alocal maximumnor alocal
minimumfor all integersn.
h(θ) = 2cos2θ +cos4θ
In Exercises 31–38, ﬁnd the extreme values on the interval.
31. f (x) = x(10−x), [−1, 3]
solution Let f (x) = x(10−x) = 10x −x
2
. Thenf

(x) = 10−2x, sothat x = 5istheonlycritical point. Asthis
critical pointisnotintheinterval [−1, 3], weonlyneedtocheckthevalueof f attheendpointstodeterminetheextreme
values. Becausef (−1) = −11andf (3) = 21, themaximumvalueof f (x) = x(10− x) ontheinterval [−1, 3] is21
whiletheminimumvalueis−11.
June 9, 2011 LTSV SSM Second Pass
Chapter Review Exercises 253
f (x) = 6x
4
−4x
6
, [−2, 2]
33. g(θ) = sin
2
θ −cosθ, [0, 2π]
solution Let g(θ) = sin
2
θ −cosθ. Then
g

(θ) = 2sinθ cosθ +sinθ = sinθ(2cosθ +1) = 0
whenθ = 0,

3
, π,

3
, 2π. Thetablebelowlists thevalueof g at eachof thecritical points andtheendpoints of the
interval [0, 2π]. Basedonthisinformation, theminimumvalueof g(θ) ontheinterval [0, 2π] is−1andthemaximum
valueis
5
4
.
θ 0 2π/3 π 4π/3 2π
g(θ) −1 5/4 1 5/4 −1
R(t ) =
t
t
2
+t +1
, [0, 3]
35. f (x) = x
2/3
−2x
1/3
, [−1, 3]
solution Let f (x) = x
2/3
− 2x
1/3
. Then f

(x) =
2
3
x
−1/3

2
3
x
−2/3
=
2
3
x
−2/3
(x
1/3
− 1), so that thecritical
pointsarex = 0andx = 1. Withf (−1) = 3, f (0) = 0, f (1) = −1andf (3) =
3

9−2
3

3≈ −0.804, it followsthat
theminimumvalueof f (x) ontheinterval [−1, 3] is−1andthemaximumvalueis3.
f (x) = 4x −tan
2
x,
_

π
4
,
π
3
_
37. f (x) = x −x
3/2
, [0, 2]
solution Wehavef

(x) = 1−
3
2
x
1/2
, sothat x =
4
9
istheonlycritical point. Then
f
_
4
9
_
=
4
27
≈ 0.148, f (0) = 0, f (2) ≈ −0.828
sothat themaximumvalueof f on[0, 2] is≈ 0.148anditsminimumvalueis≈ −0.828.
f (x) = secx −cosx,
_

π
4
,
π
4
_ 39. Findthecritical pointsandextremevaluesof
f (x) = |x −1| +|2x −6| in[0, 8].
solution Let
f (x) = |x −1| +|2x −6| =

7−3x, x < 1
5−x, 1≤ x < 3
3x −7, x ≥ 3
.
Thederivativeof f (x) isnever zerobut doesnot exist at thetransitionpointsx = 1andx = 3. Thus, thecritical points
of f arex = 1andx = 3. Withf (0) = 7, f (1) = 4, f (3) = 2andf (8) = 17, it followsthat theminimumvalueof
f (x) ontheinterval [0, 8] is2andthemaximumvalueis17.
Matchthedescriptionof f (x) withthegraphof itsderivative f

(x) inFigure1.
(a) f (x) isincreasingandconcaveup.
(b) f (x) isdecreasingandconcaveup.
(c) f (x) isincreasingandconcavedown.
In Exercises 41–46, ﬁnd the points of inﬂection.
41. y = x
3
−4x
2
+4x
solution Let y = x
3
−4x
2
+4x. Theny

= 3x
2
−8x +4andy

= 6x −8. Thus, y

> 0andy isconcaveupfor
x >
4
3
, whiley

< 0andy isconcavedownfor x <
4
3
. Hence, thereisapoint of inﬂectionat x =
4
3
.
y = x −2cosx
43. y =
x
2
x
2
+4
solution Let y =
x
2
x
2
+4
= 1−
4
x
2
+4
. Theny

=
8x
(x
2
+4)
2
and
y

=
(x
2
+4)
2
(8) −8x(2)(2x)(x
2
+4)
(x
2
+4)
4
=
8(4−3x
2
)
(x
2
+4)
3
.
Thus, y

> 0andy isconcaveupfor

2

3
< x <
2

3
,
whiley

< 0andy isconcavedownfor
|x| ≥
2

3
.
Hence, therearepointsof inﬂectionat
x = ±
2

3
.
June 9, 2011 LTSV SSM Second Pass
254 C HA P T E R 4 APPLICATIONS OF THE DERIVATIVE
y =
x
(x
2
−4)
1/3
45. f (x) =
x
3
−x
x
2
+1
solution Wehave
f

(x) =
(x
2
+1)(3x
2
−1) −(x
3
−x)(2x)
(x
2
+1)
=
x
4
+4x
2
−1
(x
2
+1)
2
f

(x) =
(x
2
+1)
2
(4x
3
+8x) −(x
4
+4x
2
−1)(2(x
2
+1)(2x))
(x
2
+1)
4
= −
4x(x
2
−3)
(x
2
+1)
3
Sincethedenominatorof f

(x) isalwayspositive, f

(x) > 0andf (x) isconcaveupforx < −

3andfor0< x <

3,
whilef

(x) < 0andf (x) isconcavedownfor −

3 < x < 0andfor

3 < x. Thustherearepointsof inﬂectionat
x = 0andat x = ±

3.
f (x) = sin2x −4cosx
In Exercises 47–56, sketch the graph, noting the transition points and asymptotic behavior.
47. y = 12x −3x
2
solution Let y = 12x − 3x
2
. Then y

= 12− 6x and y

= −6. It follows that thegraph of y = 12x − 3x
2
is
increasingfor x < 2, decreasingfor x > 2, hasalocal maximumat x = 2andisconcavedownfor all x. Because
lim
x→±∞
(12x −3x
2
) = −∞,
thegraphhasnohorizontal asymptotes. Therearealsonovertical asymptotes. Thegraphisshownbelow.
5
x
−1 1 2 3 5 4
y
10
−5
−10
y = 8x
2
−x
4
49. y = x
3
−2x
2
+3
solution Lety = x
3
−2x
2
+3. Theny

= 3x
2
−4x andy

= 6x −4. Itfollowsthatthegraphof y = x
3
−2x
2
+3
isincreasingfor x < 0andx >
4
3
, isdecreasingfor 0< x <
4
3
, hasalocal maximumat x = 0, hasalocal minimumat
x =
4
3
, isconcaveupfor x >
2
3
, isconcavedownfor x <
2
3
andhasapoint of inﬂectionat x =
2
3
. Because
lim
x→−∞
(x
3
−2x
2
+3) = −∞ and lim
x→∞
(x
3
−2x
2
+3) = ∞,
thegraphhasnohorizontal asymptotes. Therearealsonovertical asymptotes. Thegraphisshownbelow.
5
x
−1 1 2
y
10
−5
−10
y = 4x −x
3/2
51. y =
x
x
3
+1
solution Let y =
x
x
3
+1
. Then
y

=
x
3
+1−x(3x
2
)
(x
3
+1)
2
=
1−2x
3
(x
3
+1)
2
and
y

=
(x
3
+1)
2
(−6x
2
) −(1−2x
3
)(2)(x
3
+1)(3x
2
)
(x
3
+1)
4
= −
6x
2
(2−x
3
)
(x
3
+1)
3
.
It followsthat thegraphof y =
x
x
3
+1
isincreasingfor x < −1and−1< x <
3
_
1
2
, isdecreasingfor x >
3
_
1
2
, hasa
local maximumat x =
3
_
1
2
, isconcaveupfor x < −1andx >
3

2, isconcavedownfor −1< x < 0and0< x <
3

2
June 9, 2011 LTSV SSM Second Pass
Chapter Review Exercises 255
andhasapoint of inﬂectionat x =
3

2. Notethat x = −1isnot aninﬂectionpoint becausex = −1isnot inthedomain
of thefunction. Now,
lim
x→±∞
x
x
3
+1
= 0,
soy = 0isahorizontal asymptote. Moreover,
lim
x→−1−
x
x
3
+1
= ∞ and lim
x→−1+
x
x
3
+1
= −∞,
sox = −1isavertical asymptote. Thegraphisshownbelow.
2
x
−1 −2 −3 1 2 3
y
4
−2
−4
y =
x
(x
2
−4)
2/3
53. y =
1
|x +2| +1
solution Let y =
1
|x +2| +1
. Because
lim
x→±∞
1
|x +2| +1
= 0,
thegraphof thisfunctionhasahorizontal asymptoteof y = 0. Thegraphhasnovertical asymptotesas|x +2| +1≥ 1
for all x. Thegraphisshownbelow. Fromthisgraphweseethereisalocal maximumat x = −2.
0.8
x
−4 −2 −6 −8 2 4
y
1
0.4
0.2
0.6
y =
_
2−x
3
55. y =

3sinx −cosx on[0, 2π]
solution Lety =

3sinx −cosx.Theny

=

3cosx +sinx andy

= −

3sinx +cosx. Itfollowsthatthegraph
of y =

3sinx −cosx isincreasingfor 0< x < 5π/6and11π/6< x < 2π, isdecreasingfor 5π/6< x < 11π/6, has
alocal maximumatx = 5π/6, hasalocal minimumatx = 11π/6, isconcaveupfor 0< x < π/3and4π/3< x < 2π,
isconcavedownfor π/3< x < 4π/3andhaspointsof inﬂectionat x = π/3andx = 4π/3. Thegraphisshownbelow.
x
−1
1
4
5 6 2 3
y
1
y = 2x −tanx on[0, 2π]
57. Drawacurvey = f (x) for whichf

andf

havesignsasindicatedinFigure2.
x
−2 0
- + + + + - - - - +
1 3 5
FIGURE 2
solution Theﬁgurebelowdepictsacurvefor whichf

(x) andf

(x) havetherequiredsigns.
x
4 8 −4
y
June 9, 2011 LTSV SSM Second Pass
256 C HA P T E R 4 APPLICATIONS OF THE DERIVATIVE
Findthedimensionsof acylindrical canwithabottombut notopof volume4m
3
that usestheleast amount of
metal.
59. A rectangular boxof height h withsquarebaseof sideb hasvolumeV = 4m
3
. Twoof thesidefacesaremadeof
material costing\$40/m
2
. Theremainingsidescost \$20/m
2
. Whichvaluesof b andh minimizethecost of thebox?
solution Becausethevolumeof theboxis
V = b
2
h = 4 it followsthat h =
4
b
2
.
Now, thecost of theboxis
C = 40(2bh) +20(2bh) +20b
2
= 120bh +20b
2
=
480
b
+20b
2
.
Thus,
C

(b) = −
480
b
2
+40b = 0
whenb =
3

12meters. BecauseC(b) →∞asb →0+andasb →∞, itfollowsthatcostisminimizedwhenb =
3

12
metersandh =
1
3
3

12meters.
Thecornyieldonacertainfarmis
Y = −0.118x
2
+8.5x +12.9 (bushelsper acre)
wherex isthenumber of cornplantsper acre(inthousands). Assumethatcornseedcosts\$1.25(per thousandseeds)
andthat corncanbesoldfor \$1.50/bushel. Let P(x) betheproﬁt (revenueminusthecost of seeds) at plantinglevel
x.
(a) ComputeP(x
0
) for thevaluex
0
that maximizesyieldY.
(b) Findthemaximumvalueof P(x). Doesmaximumyieldleadtomaximumproﬁt?
61. A quantityN(T ) satisﬁes
dN
dt
=
2
t

8
t
2
for t ≥ 4(t indays). At whichtimeisN increasingmost rapidly?
solution N is increasingmost rapidly wheretheslopeis greatest, i.e. whereN

=
dN
dt
has alocal maximum. We
have
N

= −
2
t
2
+
16
t
3
andN

(t ) = 0for t = 8. Signanalysisshowsthat N

(t ) changessignfrompositivetonegativeat t = 8, sothat infact
t = 8isalocal maximum.
A truck gets 10 miles per gallon of diesel fuel traveling along an interstatehighway at 50 mph. This mileage
decreasesby0.15mpgfor eachmileper hour increaseabove50mph.
(a) If thetruckdriver ispaid\$30/hour anddiesel fuel costsP = \$3/gal, whichspeedv between50and70mphwill
minimizethecost of atripalongthehighway? Noticethat theactual cost dependsonthelengthof thetrip, but the
optimal speeddoesnot.
(b) Plot cost asafunctionof v (choosethelengtharbitrarily) andverifyyour answer topart (a).
(c) Doyouexpecttheoptimal speedv toincreaseor decreaseif fuel costsgodowntoP = \$2/gal?Plotthegraphs
of cost asafunctionof v for P = 2andP = 3onthesameaxisandverifyyour conclusion.
63. Findthemaximumvolumeof aright-circular coneplacedupside-downinaright-circular coneof radiusR = 3and
height H = 4asinFigure3. A coneof radiusr andheight h hasvolume
1
3
πr
2
h.
R
H
FIGURE 3
solution Letr denotetheradiusandh theheightof theupsidedowncone. Bysimilar triangles, weobtaintherelation
4−h
r
=
4
3
so h = 4
_
1−
r
3
_
andthevolumeof theupsidedownconeis
V(r) =
1
3
πr
2
h =
4
3
π
_
r
2

r
3
3
_
for 0≤ r ≤ 3. Thus,
dV
dr
=
4
3
π
_
2r −r
2
_
,
andthecritical pointsarer = 0andr = 2. BecauseV(0) = V(3) = 0and
V (2) =
4
3
π
_
4−
8
3
_
=
16
9
π,
themaximumvolumeof aright-circular coneplacedupsidedowninaright-circular coneof radius3andheight 4is
16
9
π.
June 9, 2011 LTSV SSM Second Pass
Chapter Review Exercises 257
RedoExercise63for arbitraryR andH.
65. Showthatthemaximumareaof aparallelogramADEF thatisinscribedinatriangleABC, asinFigure4, isequal
toone-half theareaof ABC.
D E
B
F
C A
FIGURE 4
solution Let θ denotethemeasureof angleBAC. Then theareaof theparallelogramis given by AD · AF sinθ.
Now, supposethat
BE/BC = x.
Then, by similar triangles, AD = (1− x)AB, AF = DE = xAC, andtheareaof theparallelogrambecomes AB ·
ACx(1−x) sinθ. Thefunctionx(1−x) achievesitsmaximumvalueof
1
4
whenx =
1
2
. Thus, themaximumareaof a
parallelograminscribedinatriangleABC is
1
4
AB · AC sinθ =
1
2
_
1
2
AB · AC sinθ
_
=
1
2
(areaof ABC) .
A box of volume8m
3
withasquaretopandbottomisconstructedout of twotypesof metal. Themetal for the
topandbottomcosts\$50/m
2
andthemetal for thesidescosts\$30/m
2
. Findthedimensionsof theboxthat minimize
total cost.
67. Let f (x) beafunctionwhosegraphdoesnot passthroughthex-axisandlet Q = (a, 0). Let P = (x
0
, f (x
0
)) be
thepointonthegraphclosesttoQ(Figure5). ProvethatPQisperpendicular tothetangentlinetothegraphof x
0
. Hint:
Findtheminimumvalueof thesquare of thedistancefrom(x, f (x)) to(a, 0).
y
x
y = f (x)
P = (x
0
, f (x
0
))
Q = (a, 0)
FIGURE 5
solution Let P = (a, 0) andlet Q = (x
0
, f (x
0
)) bethepoint onthegraphof y = f (x) closest toP. Theslopeof
thesegment joiningP andQisthen
f (x
0
)
x
0
−a
.
Now, let
q(x) =
_
(x −a)
2
+(f (x))
2
,
thedistancefromthearbitrarypoint(x, f (x)) onthegraphof y = f (x) tothepointP.As(x
0
, f (x
0
)) isthepointclosest
toP, wemust have
q

(x
0
) =
2(x
0
−a) +2f (x
0
)f

(x
0
)
_
(x
0
−a)
2
+(f (x
0
))
2
= 0.
Thus,
f

(x
0
) = −
x
0
−a
f (x
0
)
= −
_
f (x
0
)
x
0
−a
_
−1
.
Inother words, theslopeof thesegment joiningP andQisthenegativereciprocal of theslopeof thelinetangent tothe
graphof y = f (x) at x = x
0
; hence; thetwolinesareperpendicular.
Takeacircular pieceof paper of radius R, removeasector of angleθ (Figure6), andfoldtheremainingpiece
intoacone-shapedcup. Whichangleθ producesthecupof largest volume?
69. UseNewton’sMethodtoestimate
3

25tofour decimal places.
solution Let f (x) = x
3
−25anddeﬁne
x
n+1
= x
n

f (x
n
)
f

(x
n
)
= x
n

x
3
n
−25
3x
2
n
.
Withx
0
= 3, weﬁnd
June 9, 2011 LTSV SSM Second Pass
258 C HA P T E R 4 APPLICATIONS OF THE DERIVATIVE
n 1 2 3
x
n
2.925925926 2.924018982 2.924017738
Thus, tofour decimal places
3

25= 2.9240.
UseNewton’sMethodtoﬁndaroot of f (x) = x
2
−x −1tofour decimal places.
In Exercises 71–84, calculate the indeﬁnite integral.
71.
_
_
4x
3
−2x
2
_
dx
solution
_
(4x
3
−2x
2
) dx = x
4

2
3
x
3
+C.
_
x
9/4
dx
73.
_
sin(θ −8) dθ
solution
_
sin(θ −8) dθ = −cos(θ −8) +C.
_
cos(5−7θ) dθ
75.
_
(4t
−3
−12t
−4
) dt
solution
_
(4t
−3
−12t
−4
) dt = −2t
−2
+4t
−3
+C.
_
(9t
−2/3
+4t
7/3
) dt
77.
_
sec
2
x dx
solution
_
sec
2
x dx = tanx +C.
_
tan3θ sec3θ dθ
79.
_
(y +2)
4
dy
solution
_
(y +2)
4
dy =
1
5
(y +2)
5
+C.
_
3x
3
−9
x
2
dx
81.
_
(cosθ −θ) dθ
solution
_
(cosθ −θ) dθ = sinθ −
1
2
θ
2
+C.
_
sec
2
(12−25θ) dθ
83.
_
8dx
x
3
solution
_
8dx
x
3
= −
4
x
2
+C.
_
sin(4x −9) dx
In Exercises 85–90, solve the differential equation with the given initial condition.
85.
dy
dx
= 4x
3
, y(1) = 4
solution Let
dy
dx
= 4x
3
. Then
y(x) =
_
4x
3
dx = x
4
+C.
Usingtheinitial conditiony(1) = 4, weﬁndy(1) = 1
4
+C = 4, soC = 3. Thus, y(x) = x
4
+3.
dy
dt
= 3t
2
+cost , y(0) = 12
87.
dy
dx
= x
−1/2
, y(1) = 1
solution Let
dy
dx
= x
−1/2
. Then
y(x) =
_
x
−1/2
dx = 2x
1/2
+C.
Usingtheinitial conditiony(1) = 1, weﬁndy(1) = 2

1+C = 1, soC = −1. Thus, y(x) = 2x
1/2
−1.
dy
dx
= sec
2
x, y
_
π
4
_
= 2
June 9, 2011 LTSV SSM Second Pass
Chapter Review Exercises 259
89.
dy
dt
= 1+π sin3t , y(π) = π
solution Integratinggives
y(t ) =
_
(1+π sin3t ) dt = t −
π
3
cos3t +C
Usingtheinitial condition, weﬁnd
π = y(π) = π −
π
3
cos3π +C = π +
π
3
+C
sothat C = −
π
3
, andthusy(t ) = t −
π
3
(1+cos3t )
dy
dt
= cos3πt +sin4πt , y
_
1
3
_
= 0
91. Findf (t ) if f

(t ) = 1−2t , f (0) = 2, andf

(0) = −1.
solution Supposef

(t ) = 1−2t . Then
f

(t ) =
_
f

(t ) dt =
_
(1−2t ) dt = t −t
2
+C.
Usingtheinitial conditionf

(0) = −1, weﬁndf

(0) = 0−0
2
+C = −1, soC = −1. Thus, f

(t ) = t −t
2
−1. Now,
f (t ) =
_
f

(t ) dt =
_
(t −t
2
−1) dt =
1
2
t
2

1
3
t
3
−t +C.
Usingtheinitial conditionf (0) = 2, weﬁndf (0) =
1
2
0
2

1
3
0
3
−0+C = 2, soC = 2. Thus,
f (t ) =
1
2
t
2

1
3
t
3
−t +2.
At timet = 0, adriver begins deceleratingat aconstant rateof −10m/s
2
andcomes to ahalt after traveling
500m. Findthevelocityat t = 0.
June 9, 2011 LTSV SSM Second Pass
5 THE INTEGRAL
5.1 Approximating and Computing Area
Preliminary Questions
1. What aretheright andleft endpointsif [2, 5] isdividedintosixsubintervals?
solution If theinterval [2, 5] is dividedintosix subintervals, thelengthof eachsubinterval is
5−2
6
=
1
2
. Theright
endpointsof thesubintervalsarethen
5
2
, 3,
7
2
, 4,
9
2
, 5, whiletheleft endpointsare2,
5
2
, 3,
7
2
, 4,
9
2
.
2. Theinterval [1, 5] isdividedintoeight subintervals.
(a) What istheleft endpoint of thelast subinterval?
(b) What aretheright endpointsof theﬁrst twosubintervals?
solution Notethat eachof the8subintervalshaslength
5−1
8
=
1
2
.
(a) Theleft endpoint of thelast subinterval is5−
1
2
=
9
2
.
(b) Theright endpointsof theﬁrst twosubintervalsare1+
1
2
=
3
2
and1+2
_
1
2
_
= 2.
3. Whichof thefollowingpairsof sumsarenot equal?
(a)
4

i=1
i,
4

=1
(b)
4

j=1
j
2
,
5

k=2
k
2
(c)
4

j=1
j,
5

i=2
(i −1) (d)
4

i=1
i(i +1),
5

j=2
(j −1)j
solution
(a) Onlythenameof theindexvariablehasbeenchanged, sothesetwosumsare thesame.
(b) Thesetwosumsarenot thesame; thesecondsquaresthenumberstwothroughﬁvewhiletheﬁrstsquaresthenumbers
onethroughfour.
(c) Thesetwosumsare thesame. Notethat wheni rangesfromtwothroughﬁve, theexpressioni −1rangesfromone
throughfour.
(d) Thesetwosumsare thesame. Bothsumsare1· 2+2· 3+3· 4+4· 5.
4. Explain:
100

j=1
j =
100

j=0
j but
100

j=1
1isnot equal to
100

j=0
1.
solution Theﬁrst terminthesum

100
j=0
j isequal tozero, soit maybedropped. Morespeciﬁcally,
100

j=0
j = 0+
100

j=1
j =
100

j=1
j.
Ontheother hand, theﬁrst termin

100
j=0
1isnot zero, sothistermcannot bedropped. Inparticular,
100

j=0
1= 1+
100

j=1
1=
100

j=1
1.
5. ExplainwhyL
100
≥ R
100
for f (x) = x
−2
on[3, 7].
solution On[3, 7], thefunctionf (x) = x
−2
isadecreasingfunction; hence, foranysubinterval of [3, 7], thefunction
valueat theleft endpoint islarger thanthefunctionvalueat theright endpoint. Consequently, L
100
must belarger than
R
100
.
260
June 9, 2011 LTSV SSM Second Pass
S E C T I ON 5.1 Approximating and Computing Area 261
Exercises
1. Figure15showsthevelocity of anobject over a3-mininterval. Determinethedistancetraveledover theintervals
[0, 3] and[1, 2.5] (remember toconvert fromkm/htokm/min).
3
min
km/h
2 1
20
30
10
FIGURE 15
solution Thedistancetraveledbytheobject canbedeterminedbycalculatingtheareaunderneaththevelocitygraph
over thespeciﬁedinterval. Duringtheinterval [0, 3], theobject travels
_
10
60
__
1
2
_
+
_
25
60
_
(1) +
_
15
60
__
1
2
_
+
_
20
60
_
(1) =
23
24
≈ 0.96km.
Duringtheinterval [1, 2.5], it travels
_
25
60
__
1
2
_
+
_
15
60
__
1
2
_
+
_
20
60
__
1
2
_
=
1
2
= 0.5km.
Anostrich(Figure16) runswithvelocity20km/hfor 2minutes, 12km/hfor 3minutes, and40km/hfor another
minute. Computethetotal distancetraveledandindicatewithagraphhowthisquantitycanbeinterpretedasanarea.
3. A rainstormhit Portland, Maine, inOctober 1996, resultinginrecordrainfall. Therainfall rateR(t ) onOctober 21
isrecorded, incentimetersper hour, inthefollowingtable, wheret isthenumber of hourssincemidnight. Computethe
total rainfall duringthis24-hour periodandindicateonagraphhowthisquantitycanbeinterpretedasanarea.
t (h) 0–2 2–4 4–9 9–12 12–20 20–24
R(t ) (cm) 0.5 0.3 1.0 2.5 1.5 0.6
solution Over eachinterval, thetotal rainfall isthetimeinterval inhourstimestherainfall incentimetersper hour.
Thus
R = 2(0.5) +2(0.3) +5(1.0) +3(2.5) +8(1.5) +4(0.6) = 28.5cm.
Theﬁgurebelowisagraphof therainfall asafunctionof time. Theareaof theshadedregionrepresentsthetotal rainfall.
0.5
1.0
1.5
2.0
2.5
5 10 15 20 25
y
x
Thevelocityof anobject isv(t ) = 12t m/s. UseEq. (2) andgeometrytoﬁndthedistancetraveledover thetime
intervals[0, 2] and[2, 5].
5. ComputeR
5
andL
5
over [0, 1] usingthefollowingvalues.
x 0 0.2 0.4 0.6 0.8 1
f (x) 50 48 46 44 42 40
solution x =
1−0
5
= 0.2. Thus,
L
5
= 0.2(50+48+46+44+42) = 0.2(230) = 46,
and
R
5
= 0.2(48+46+44+42+40) = 0.2(220) = 44.
Theaverageis
46+44
2
= 45.
ThisestimateisfrequentlyreferredtoastheTrapezoidal Approximation.
June 9, 2011 LTSV SSM Second Pass
262 C HA P T E R 5 THE INTEGRAL
ComputeR
6
, L
6
, andM
3
toestimatethedistancetraveledover [0, 3] if thevelocityat half-secondintervalsisas
follows:
t (s) 0 0.5 1 1.5 2 2.5 3
v (m/s) 0 12 18 25 20 14 20
7. Let f (x) = 2x +3.
(a) ComputeR
6
andL
6
over [0, 3].
(b) Usegeometrytoﬁndtheexact areaA andcomputetheerrors|A −R
6
| and|A −L
6
| intheapproximations.
solution Let f (x) = 2x +3on[0, 3].
(a) Wepartition[0, 3] into 6equally-spacedsubintervals. Theleft endpoints of thesubintervals are
_
0,
1
2
, 1,
3
2
, 2,
5
2
_
whereastheright endpointsare
_
1
2
, 1,
3
2
, 2,
5
2
, 3
_
.
• Let a = 0, b = 3, n = 6, x = (b −a) /n =
1
2
, andx
k
= a +kx, k = 0, 1, . . . , 5(left endpoints). Then
L
6
=
5

k=0
f (x
k
)x = x
5

k=0
f (x
k
) =
1
2
(3+4+5+6+7+8) = 16.5.
• Withx
k
= a +kx, k = 1, 2, . . . , 6(right endpoints), wehave
R
6
=
6

k=1
f (x
k
)x = x
6

k=1
f (x
k
) =
1
2
(4+5+6+7+8+9) = 19.5.
(b) Viageometry (seeﬁgurebelow), theexact areaisA =
1
2
(3) (6) + 3
2
= 18. Thus, L
6
underestimatesthetruearea
(L
6
−A = −1.5), whileR
6
overestimatesthetruearea(R
6
−A = +1.5).
0.5 1 1.5 2 2.5 3
3
6
9
x
y
Repeat Exercise7for f (x) = 20−3x over [2, 4].
9. CalculateR
3
andL
3
for f (x) = x
2
−x +4 over [1, 4]
Thensketchthegraphof f andtherectangles that makeupeachapproximation. Is theareaunder thegraphlarger or
smaller thanR
3
? Isit larger or smaller thanL
3
?
solution Let f (x) = x
2
−x +4andset a = 1, b = 4, n = 3, x = (b −a) /n = (4−1) /3= 1.
(a) Let x
k
= a +kx, k = 0, 1, 2, 3.
• Selectingtheleft endpointsof thesubintervals, x
k
, k = 0, 1, 2, or {1, 2, 3}, wehave
L
3
=
2

k=0
f (x
k
)x = x
2

k=0
f (x
k
) = (1) (4+6+10) = 20.
• Selectingtheright endpointsof thesubintervals, x
k
, k = 1, 2, 3, or {2, 3, 4}, wehave
R
3
=
3

k=1
f (x
k
)x = x
3

k=1
f (x
k
) = (1) (6+10+16) = 32.
(b) Hereareﬁguresof thethreerectanglesthatapproximatetheareaunderthecurvef (x) overtheinterval [1, 4]. Clearly,
theareaunder thegraphislarger thanL
3
but smaller thanR
3
.
4
1.0 1.5 2.0 2.5 3.0 3.5
6
8
10
12
14
y
x
L
3
4
1.0 1.5 2.0 2.5 3.0 3.5
6
8
10
12
14
y
x
R
3
June 9, 2011 LTSV SSM Second Pass
S E C T I ON 5.1 Approximating and Computing Area 263
Let f (x) =
_
x
2
+1andx =
1
3
. Sketchthegraphof f (x) anddrawtheright-endpoint rectangleswhosearea
isrepresentedbythesum
6

i=1
f (1+ix)x.
11. EstimateR
3
, M
3
, andL
6
over [0, 1.5] for thefunctioninFigure17.
1
2
3
4
5
x
y
0.5 1 1.5
FIGURE 17
solution Let f (x) on [0,
3
2
] begiven by Figure17. For n = 3, x = (
3
2
− 0)/3 =
1
2
, {x
k
}
3
k=0
=
_
0,
1
2
, 1,
3
2
_
.
Therefore
R
3
=
1
2
3

k=1
f (x
k
) =
1
2
(2+1+2) = 2.5,
M
3
=
1
2
6

k=1
f
_
x
k

1
2
x
_
=
1
2
(3.25+1.25+1.25) = 2.875.
For n = 6, x = (
3
2
−0)/6=
1
4
, {x
k
}
6
k=0
=
_
0,
1
4
,
1
2
,
3
4
, 1,
5
4
,
3
2
_
. Therefore
L
6
=
1
4
5

k=0
f (x
k
) =
1
4
(5+3.25+2+1.25+1+1.25) = 3.4375.
Calculatetheareaof theshadedrectanglesinFigure18. Whichapproximationdotheserectanglesrepresent?
In Exercises 13–20, calculate the approximation for the given function and interval.
13. R
3
, f (x) = 7−x, [3, 5]
solution Let f (x) = 7−x on[3, 5]. For n = 3, x = (5−3)/3=
2
3
, and{x
k
}
3
k=0
=
_
3,
11
3
,
13
3
, 5
_
. Therefore
R
3
=
2
3
3

k=1
(7−x
k
)
=
2
3
_
10
3
+
8
3
+2
_
=
2
3
(8) =
16
3
.
L
6
, f (x) =

6x +2, [1, 3]
15. M
6
, f (x) = 4x +3, [5, 8]
solution Let f (x) = 4x + 3on[5, 8]. For n = 6, x = (8− 5)/6 =
1
2
, and{x

k
}
5
k=0
= {5.25, 5.75, 6.25, 6.75,
7.25, 7.75}. Therefore,
M
6
=
1
2
5

k=0
_
4x

k
+3
_
=
1
2
(24+26+28+30+32+34)
=
1
2
(174) = 87.
R
5
, f (x) = x
2
+x, [−1, 1]
17. L
6
, f (x) = x
2
+3|x|, [−2, 1]
solution Let f (x) = x
2
+ 3|x| on[−2, 1]. For n = 6, x = (1− (−2))/6 =
1
2
, and{x
k
}
6
k=0
= {−2, −1.5, −1,
−0.5, 0, 0.5, 1}. Therefore
L
6
=
1
2
5

k=0
(x
2
k
+3|x
k
|) =
1
2
(10+6.75+4+1.75+0+1.75) = 12.125.
M
4
, f (x) =

x, [3, 5]
19. L
4
, f (x) = cos
2
x,
_
π
6
,
π
2
_
solution Let f (x) = cos
2
x on[
π
6
,
π
2
]. For n = 4,
x =
(π/2−π/6)
4
=
π
12
and {x
k
}
4
k=0
=
_
π
6
,
π
4
,
π
3
,

12
,
π
2
_
.
June 9, 2011 LTSV SSM Second Pass
264 C HA P T E R 5 THE INTEGRAL
Therefore
L
4
=
π
12
3

k=0
cos
2
x
k
≈ 0.410236.
M
4
, f (x) =
1
x
2
+1
, [1, 5]
In Exercises 21–26, write the sum in summation notation.
21. 4
7
+5
7
+6
7
+7
7
+8
7
solution Theﬁrst termis4
7
, andthelast termis8
7
, soit seemsthekthtermisk
7
. Therefore, thesumis:
8

k=4
k
7
.
(2
2
+2) +(3
2
+3) +(4
2
+4) +(5
2
+5)
23. (2
2
+2) +(2
3
+2) +(2
4
+2) +(2
5
+2)
solution Theﬁrst termis2
2
+2, andthelast termis2
5
+2, soit seemsthesumlimitsare2and5, andthekthterm
is2
k
+2. Therefore, thesumis:
5

k=2
(2
k
+2).
_
1+1
3
+
_
2+2
3
+· · · +
_
n +n
3
25.
1
2· 3
+
2
3· 4
+· · · +
n
(n +1)(n +2)
solution Theﬁrst summandis
1
(1+1)·(1+2)
. Thisshowsus
1
2· 3
+
2
3· 4
+· · · +
n
(n +1)(n +2)
=
n

i=1
i
(i +1)(i +2)
.
sin(π) +sin
_
π
2
_
+sin
_
π
3
_
+· · · +sin
_
π
n +1
_ 27. Calculatethesums:
(a)
5

i=1
9 (b)
5

i=0
4 (c)
4

k=2
k
3
solution
(a)
5

i=1
9= 9+9+9+9+9= 45. Alternatively,
5

i=1
9= 9
5

i=1
1= (9)(5) = 45.
(b)
5

i=0
4= 4+4+4+4+4+4= 24. Alternatively,
5

i=0
4= 4
5

i=0
= (4)(6) = 24.
(c)
4

k=2
k
3
= 2
3
+3
3
+4
3
= 99. Alternatively,
4

k=2
k
3
=

4

k=1
k
3

1

k=1
k
3

=
_
4
4
4
+
4
3
2
+
4
2
4
_

_
1
4
4
+
1
3
2
+
1
2
4
_
= 99.
Calculatethesums:
(a)
4

j=3
sin
_
j
π
2
_
(b)
5

k=3
1
k −1
(c)
2

j=0
3
j−1
29. Let b
1
= 4, b
2
= 1, b
3
= 2, andb
4
= −4. Calculate:
(a)
4

i=2
b
i
(b)
2

j=1
(2
b
j
−b
j
) (c)
3

k=1
kb
k
solution
(a)
4

i=2
b
i
= b
2
+b
3
+b
4
= 1+2+(−4) = −1.
(b)
2

j=1
_
2
b
j
−b
j
_
= (2
4
−4) +(2
1
−1) = 13.
(c)
3

k=1
kb
k
= 1(4) +2(1) +3(2) = 12.
June 9, 2011 LTSV SSM Second Pass
S E C T I ON 5.1 Approximating and Computing Area 265
Assumethat a
1
= −5,
10

i=1
a
i
= 20, and
10

i=1
b
i
= 7. Calculate:
(a)
10

i=1
(4a
i
+3) (b)
10

i=2
a
i
(c)
10

i=1
(2a
i
−3b
i
)
31. Calculate
200

j=101
j. Hint: Writeasadifferenceof twosumsanduseformula(3).
solution
200

j=101
j =
200

j=1
j −
100

j=1
j =
_
200
2
2
+
200
2
_

_
100
2
2
+
100
2
_
= 20,100−5050= 15,050.
Calculate
30

j=1
(2j +1)
2
. Hint: Expandanduseformulas(3)–(4).
In Exercises 33–40, use linearity and formulas (3)–(5) to rewrite and evaluate the sums.
33.
20

j=1
8j
3
solution
20

j=1
8j
3
= 8
20

j=1
j
3
= 8
_
20
4
4
+
20
3
2
+
20
2
4
_
= 8(44,100) = 352,800.
30

k=1
(4k −3)
35.
150

n=51
n
2
solution
150

n=51
n
2
=
150

n=1
n
2

50

n=1
n
2
=
_
150
3
3
+
150
2
2
+
150
6
_

_
50
3
3
+
50
2
2
+
50
6
_
= 1,136,275−42,925= 1,093,350.
200

k=101
k
3
37.
50

j=0
j (j −1)
solution
50

j=0
j (j −1) =
50

j=0
(j
2
−j) =
50

j=0
j
2

50

j=0
j
=
_
50
3
3
+
50
2
2
+
50
6
_

_
50
2
2
+
50
2
_
=
50
3
3

50
3
=
124,950
3
= 41,650.
Thepower sumformulaisusablebecause
50

j=0
j (j −1) =
50

j=1
j (j −1).
30

j=2
_
6j +
4j
2
3
_
39.
30

m=1
(4−m)
3
solution
30

m=1
(4−m)
3
=
30

m=1
(64−48m+12m
2
−m
3
)
= 64
30

m=1
1−48
30

m=1
m+12
30

m=1
m
2

30

m=1
m
3
= 64(30) −48
(30)(31)
2
+12
_
30
3
3
+
30
2
2
+
30
6
_

_
30
4
4
+
30
3
2
+
30
2
4
_
= 1920−22,320+113,460−216,225= −123,165.
June 9, 2011 LTSV SSM Second Pass
266 C HA P T E R 5 THE INTEGRAL
20

m=1
_
5+
3m
2
_
2
In Exercises 41–44, use formulas (3)–(5) to evaluate the limit.
41. lim
N→∞
N

i=1
i
N
2
solution Let s
N
=
N

i=1
i
N
2
. Then,
s
N
=
N

i=1
i
N
2
=
1
N
2
N

i=1
i =
1
N
2
_
N
2
2
+
N
2
_
=
1
2
+
1
2N
.
Therefore, lim
N→∞
s
N
=
1
2
.
lim
N→∞
N

j=1
j
3
N
4
43. lim
N→∞
N

i=1
i
2
−i +1
N
3
solution Let s
N
=
N

i=1
i
2
−i +1
N
3
. Then
s
N
=
N

i=1
i
2
−i +1
N
3
=
1
N
3

N

i=1
i
2

N

i=1
i

+

N

i=1
1

=
1
N
3
__
N
3
3
+
N
2
2
+
N
6
_

_
N
2
2
+
N
2
_
+N
_
=
1
3
+
2
3N
2
.
Therefore, lim
N→∞
s
N
=
1
3
.
lim
N→∞
N

i=1
_
i
3
N
4

20
N
_
In Exercises 45–50, calculate the limit for the given function and interval. Verify your answer by using geometry.
45. lim
N→∞
R
N
, f (x) = 9x, [0, 2]
solution Let f (x) = 9x on [0, 2]. Let N beapositiveinteger and set a = 0, b = 2, and x = (b − a)/N =
(2− 0)/N = 2/N. Also, let x
k
= a + kx = 2k/N, k = 1, 2, . . . , N betheright endpointsof theN subintervalsof
[0, 2]. Then
R
N
= x
N

k=1
f (x
k
) =
2
N
N

k=1
9
_
2k
N
_
=
36
N
2
N

k=1
k =
36
N
2
_
N
2
2
+
N
2
_
= 18+
18
N
.
Theareaunder thegraphis
lim
N→∞
R
N
= lim
N→∞
_
18+
18
N
_
= 18.
Theregionunder thegraphisatrianglewithbase2andheight 18. Theareaof theregionisthen
1
2
(2)(18) = 18, which
agreeswiththevalueobtainedfromthelimit of theright-endpoint approximations.
lim
N→∞
R
N
, f (x) = 3x +6, [1, 4]
47. lim
N→∞
L
N
, f (x) =
1
2
x +2, [0, 4]
solution Let f (x) =
1
2
x +2on[0, 4]. Let N > 0beaninteger, andset a = 0, b = 4, andx = (4−0)/N =
4
N
.
Also, let x
k
= 0+kx =
4k
N
, k = 0, 1, . . . , N −1betheleft endpointsof theN subintervals. Then
L
N
= x
N−1

k=0
f (x
k
) =
4
N
N−1

k=0
_
1
2
_
4k
N
_
+2
_
=
8
N
N−1

k=0
1+
8
N
2
N−1

k=0
k
= 8+
8
N
2
_
(N −1)
2
2
+
N −1
2
_
= 12−
4
N
.
Theareaunder thegraphis
lim
N→∞
L
N
= 12.
Theregionunder thecurveover [0, 4] is atrapezoidwithbasewidth4andheights 2and4. Fromthis, weget that the
areais
1
2
(4)(2+4) = 12, whichagreeswiththeanswer obtainedfromthelimit of theleft-endpoint approximations.
June 9, 2011 LTSV SSM Second Pass
S E C T I ON 5.1 Approximating and Computing Area 267
lim
N→∞
L
N
, f (x) = 4x −2, [1, 3]
49. lim
N→∞
M
N
, f (x) = x, [0, 2]
solution Let f (x) = x on[0, 2]. Let N > 0beaninteger andset a = 0, b = 2, andx = (b − a)/N =
2
N
. Also,
let x

k
= 0+(k −
1
2
)x =
2k−1
N
, k = 1, 2, . . . N, bethemidpointsof theN subintervalsof [0, 2]. Then
M
N
= x
N

k=1
f (x

k
) =
2
N
N

k=1
2k −1
N
=
2
N
2
N

k=1
(2k −1)
=
2
N
2

2
N

k=1
k −N

=
4
N
2
_
N
2
2
+
N
2
_

2
N
= 2.
Theareaunder thecurveover [0, 2] is
lim
N→∞
M
N
= 2.
Theregionunder thecurveover [0, 2] isatrianglewithbaseandheight 2, andthusarea2, whichagreeswiththeanswer
obtainedfromthelimit of themidpoint approximations.
lim
N→∞
M
N
, f (x) = 12−4x, [2, 6]
51. Show, for f (x) = 3x
2
+4x over [0, 2], that
R
N
=
2
N
N

j=1
_
24j
2
N
2
+
16j
N
_
Thenevaluate lim
N→∞
R
N
.
solution Let f (x) = 3x
2
+4x on[0, 2]. Let N beapositiveinteger andset a = 0, b = 2, andx = (b −a)/N =
(2− 0)/N = 2/N. Also, let x
j
= a + jx = 2j/N, j = 1, 2, . . . , N betheright endpointsof theN subintervalsof
[0, 3]. Then
R
N
= x
N

j=1
f (x
j
) =
2
N
N

j=1
_
3
_
2j
N
_
2
+4
2j
N
_
=
2
N
N

j=1
_
12j
2
N
2
+
8j
N
_
Continuing, weﬁnd
R
N
=
24
N
3
N

j=1
j
2
+
16
N
2
N

j=1
j
=
24
N
3
_
N
3
3
+
N
2
2
+
N
6
_
+
16
N
2
_
N
2
2
+
N
2
_
= 16+
20
N
+
4
N
2
Thus,
lim
N→∞
R
N
= lim
N→∞
_
16+
20
N
+
4
N
2
_
= 16.
Show, for f (x) = 3x
3
−x
2
over [1, 5], that
R
N
=
4
N
N

j=1
_
192j
3
N
3
+
128j
2
N
2
+
28j
N
+2
_
Thenevaluate lim
N→∞
R
N
.
In Exercises 53–60, ﬁnd a formula for R
N
and compute the area under the graph as a limit.
53. f (x) = x
2
, [0, 1]
solution Let f (x) = x
2
ontheinterval [0, 1]. Thenx =
1−0
N
=
1
N
anda = 0. Hence,
R
N
= x
N

j=1
f (0+jx) =
1
N
N

j=1
j
2
1
N
2
=
1
N
3
_
N
3
3
+
N
2
2
+
N
6
_
=
1
3
+
1
2N
+
1
6N
2
and
lim
N→∞
R
N
= lim
N→∞
_
1
3
+
1
2N
+
1
6N
2
_
=
1
3
.
June 9, 2011 LTSV SSM Second Pass
268 C HA P T E R 5 THE INTEGRAL
f (x) = x
2
, [−1, 5]
55. f (x) = 6x
2
−4, [2, 5]
solution Let f (x) = 6x
2
−4ontheinterval [2, 5]. Thenx =
5−2
N
=
3
N
anda = 2. Hence,
R
N
= x
N

j=1
f (2+jx) =
3
N
N

j=1
_
6
_
2+
3j
N
_
2
−4
_
=
3
N
N

j=1
_
20+
72j
N
+
54j
2
N
2
_
= 60+
216
N
2
N

j=1
j +
162
N
3
N

j=1
j
2
= 60+
216
N
2
_
N
2
2
+
N
2
_
+
162
N
3
_
N
3
3
+
N
2
2
+
N
6
_
= 222+
189
N
+
27
N
2
and
lim
N→∞
R
N
= lim
N→∞
_
222+
189
N
+
27
N
2
_
= 222.
f (x) = x
2
+7x, [6, 11]
57. f (x) = x
3
−x, [0, 2]
solution Let f (x) = x
3
−x ontheinterval [0, 2]. Thenx =
2−0
N
=
2
N
anda = 0. Hence,
R
N
= x
N

j=1
f (0+jx) =
2
N
N

j=1
_
_
2j
N
_
3

2j
N
_
=
2
N
N

j=1
_
8j
3
N
3

2j
N
_
=
16
N
4
N

j=1
j
3

4
N
2
N

j=1
j
=
16
N
4
_
N
4
4
+
N
3
2
+
N
2
2
_

4
N
2
_
N
2
2
+
N
2
_
= 2+
6
N
+
8
N
2
and
lim
N→∞
R
N
= lim
N→∞
_
2+
6
N
+
8
N
2
_
= 2.
f (x) = 2x
3
+x
2
, [−2, 2]
59. f (x) = 2x +1, [a, b] (a, b constantswitha < b)
solution Let f (x) = 2x +1ontheinterval [a, b]. Thenx =
b −a
N
. Hence,
R
N
= x
N

j=1
f (a +jx) =
(b −a)
N
N

j=1
_
2
_
a +j
(b −a)
N
_
+1
_
=
(b −a)
N
(2a +1)
N

j=1
1+
2(b −a)
2
N
2
N

j=1
j
=
(b −a)
N
(2a +1)N +
2(b −a)
2
N
2
_
N
2
2
+
N
2
_
= (b −a)(2a +1) +(b −a)
2
+
(b −a)
2
N
and
lim
N→∞
R
N
= lim
N→∞
_
(b −a)(2a +1) +(b −a)
2
+
(b −a)
2
N
_
= (b −a)(2a +1) +(b −a)
2
= (b
2
+b) −(a
2
+a).
June 9, 2011 LTSV SSM Second Pass
S E C T I ON 5.1 Approximating and Computing Area 269
f (x) = x
2
, [a, b] (a, b constantswitha < b)
In Exercises 61–64, describe the area represented by the limits.
61. lim
N→∞
1
N
N

j=1
_
j
N
_
4
solution Thelimit
lim
N→∞
R
N
= lim
N→∞
1
N
N

j=1
_
j
N
_
4
representstheareabetweenthegraphof f (x) = x
4
andthex-axisover theinterval [0, 1].
lim
N→∞
3
N
N

j=1
_
2+
3j
N
_
4
63. lim
N→∞
5
N
N−1

j=0
_
−2+
5j
N
_
solution Thelimit
lim
N→∞
L
N
= lim
N→∞
5
N
N−1

j=0
_
−2+
5j
N
_
representstheareabetweenthegraphof y = x andthex-axisover theinterval [−2, 3].
lim
N→∞
π
2N
N

j=1
sin
_
π
3

π
4N
+

2N
_
In Exercises 65–70, express the area under the graph as a limit using the approximation indicated (in summation notation),
but do not evaluate.
65. R
N
, f (x) = sinx over [0, π]
solution Let f (x) = sinx over [0, π] andset a = 0, b = π, andx = (b −a) /N = π/N. Then
R
N
= x
N

k=1
f (x
k
) =
π
N
N

k=1
sin
_

N
_
.
Hence
lim
N→∞
R
N
= lim
N→∞
π
N
N

k=1
sin
_

N
_
istheareabetweenthegraphof f (x) = sinx andthex-axisover [0, π].
R
N
, f (x) = x
−1
over [1, 7]
67. L
N
, f (x) =

2x +1over [7, 11]
solution Let f (x) =

2x +1over theinterval [7, 11]. Thenx =
11−7
N
=
4
N
anda = 7. Hence,
L
N
= x
N−1

j=0
f (7+jx) =
4
N
N−1

j=0
_
2(7+j
4
N
) +1
and
lim
N→∞
L
N
= lim
N→∞
4
N
N−1

j=0
_
15+
8j
N
istheareabetweenthegraphof f (x) =

2x +1andthex-axisover [7, 11].
L
N
, f (x) = cosx over
_
π
8
,
π
4
_
69. M
N
, f (x) = tanx over
_
1
2
, 1
_
solution Let f (x) = tanx over theinterval [
1
2
, 1]. Thenx =
1−
1
2
N
=
1
2N
anda =
1
2
. Hence
M
N
= x
N

j=1
f
_
1
2
+
_
j −
1
2
_
x
_
=
1
2N
N

j=1
tan
_
1
2
+
1
2N
_
j −
1
2
__
andso
lim
N→∞
M
N
= lim
N→∞
1
2N
N

j=1
tan
_
1
2
+
1
2N
_
j −
1
2
__
istheareabetweenthegraphof f (x) = tanx andthex-axisover [
1
2
, 1].
June 9, 2011 LTSV SSM Second Pass
270 C HA P T E R 5 THE INTEGRAL
M
N
, f (x) = x
−2
over [3, 5]
71. Evaluate lim
N→∞
1
N
N

j=1
_
1−
_
j
N
_
2
byinterpretingit astheareaof part of afamiliar geometricﬁgure.
solution Thelimit
lim
N→∞
R
N
= lim
N→∞
1
N
N

j=1
_
1−
_
j
N
_
2
representstheareabetweenthegraphof y = f (x) =
_
1−x
2
andthex-axisover theinterval [0, 1]. Thisistheportion
of thecircular diskx
2
+y
2
≤ 1that liesintheﬁrst quadrant. Accordingly, itsareais
1
4
π (1)
2
=
π
4
.
In Exercises 72–74, let f (x) = x
2
and let R
N
, L
N
, and M
N
be the approximations for the interval [0, 1].
Showthat R
N
=
1
3
+
1
2N
+
1
6N
2
. Interpret thequantity
1
2N
+
1
6N
2
astheareaof aregion.
73. Showthat
L
N
=
1
3

1
2N
+
1
6N
2
, M
N
=
1
3

1
12N
2
ThenrankthethreeapproximationsR
N
, L
N
, andM
N
inorder of increasingaccuracy(useExercise72).
solution Let f (x) = x
2
on[0, 1]. Let N beapositiveinteger andset a = 0, b = 1, andx = (b −a) /N = 1/N.
Let x
k
= a +kx = k/N, k = 0, 1, . . . , N andlet x

k
= a +(k +
1
2
)x = (k +
1
2
)/N, k = 0, 1, . . . , N −1. Then
L
N
= x
N−1

k=0
f (x
k
) =
1
N
N−1

k=0
_
k
N
_
2
=
1
N
3
N−1

k=1
k
2
=
1
N
3
_
(N −1)
3
3
+
(N −1)
2
2
+
N −1
6
_
=
1
3

1
2N
+
1
6N
2
M
N
= x
N−1

k=0
f (x

k
) =
1
N
N−1

k=0
_
k +
1
2
N
_
2
=
1
N
3
N−1

k=0
_
k
2
+k +
1
4
_
=
1
N
3

N−1

k=1
k
2

+

N−1

k=1
k

+
1
4

N−1

k=0
1

=
1
N
3
__
(N −1)
3
3
+
(N −1)
2
2
+
N −1
6
_
+
_
(N −1)
2
2
+
N −1
2
_
+
1
4
N
_
=
1
3

1
12N
2
Theerror of R
N
isgivenby
1
2N
+
1
6N
2
, theerror of L
N
isgivenby−
1
2N
+
1
6N
2
andtheerror of M
N
isgivenby

1
12N
2
. Of thethreeapproximations, R
N
istheleast accurate, thenL
N
andﬁnallyM
N
isthemost accurate.
For eachof R
N
, L
N
, andM
N
, ﬁndthesmallest integer N for whichtheerror islessthan0.001.
In Exercises 75–80, use the Graphical Insight on page 291 to obtain bounds on the area.
75. LetA betheareaunder f (x) =

x over [0, 1]. Provethat0.51≤ A ≤ 0.77bycomputingR
4
andL
4
. Explainyour
reasoning.
solution For n = 4, x =
1−0
4
=
1
4
and{x
i
}
4
i=0
= {0+ix} = {0,
1
4
,
1
2
,
3
4
, 1}. Therefore,
R
4
= x
4

i=1
f (x
i
) =
1
4
_
1
2
+

2
2
+

3
2
+1
_
≈ 0.768
L
4
= x
3

i=0
f (x
i
) =
1
4
_
0+
1
2
+

2
2
+

3
2
_
≈ 0.518.
Intheplot below, youcanseetherectangleswhoseareaisrepresentedbyL
4
under thegraphandthetopof thosewhose
areaisrepresentedbyR
4
abovethegraph. TheareaA under thecurveissomewherebetweenL
4
andR
4
, so
0.518≤ A ≤ 0.768.
June 9, 2011 LTSV SSM Second Pass
S E C T I ON 5.1 Approximating and Computing Area 271
L
4
, R
4
andthegraphof f (x).
UseR
5
andL
5
toshowthat theareaA under y = x
−2
over [10, 13] satisﬁes0.0218≤ A ≤ 0.0244.
77. UseR
4
andL
4
toshowthat theareaA under thegraphof y = sinx over
_
0,
π
2
_
satisﬁes0.79≤ A ≤ 1.19.
solution Let f (x) = sinx. f (x) isincreasingover theinterval [0, π/2], sotheInsight onpage291applies, which
indicatesthat L
4
≤ A ≤ R
4
. For n = 4, x =
π/2−0
4
=
π
8
and{x
i
}
4
i=0
= {0+ ix}
4
i=0
= {0,
π
8
,
π
4
,

8
,
π
2
}. From
this,
L
4
=
π
8
3

i=0
f (x
i
) ≈ 0.79, R
4
=
π
8
4

i=1
f (x
i
) ≈ 1.18.
HenceA isbetween0.79and1.19.
Left andRight endpoint approximationstoA.
Showthat theareaA under f (x) = x
−1
over [1, 8] satisﬁes
1
2
+
1
3
+
1
4
+
1
5
+
1
6
+
1
7
+
1
8
≤ A ≤ 1+
1
2
+
1
3
+
1
4
+
1
5
+
1
6
+
1
7
79. Showthat theareaA under y = x
1/4
over [0, 1] satisﬁesL
N
≤ A ≤ R
N
for all N. Useacomputer algebra
systemtocalculateL
N
andR
N
for N = 100and200, anddetermineA totwodecimal places.
solution On[0, 1], f (x) = x
1/4
isanincreasingfunction; therefore, L
N
≤ A ≤ R
N
for all N. Weﬁnd
L
100
= 0.793988 and R
100
= 0.80399,
while
L
200
= 0.797074 and R
200
= 0.802075.
Thus, A = 0.80totwodecimal places.
Showthat theareaA under y = 4/(x
2
+ 1) over [0, 1] satisﬁesR
N
≤ A ≤ L
N
for all N. DetermineA toat
least threedecimal placesusingacomputer algebrasystem. Canyouguesstheexact valueof A?
Further Insights and Challenges
81. Althoughtheaccuracyof R
N
generallyimprovesasN increases, thisneednot betruefor small valuesof N. Draw
thegraphof apositivecontinuousfunctionf (x) onaninterval suchthat R
1
iscloser thanR
2
totheexact areaunder the
graph. Cansuchafunctionbemonotonic?
solution Let δ beasmall positivenumber lessthan
1
4
. (Intheﬁguresbelow, δ =
1
10
. But imagineδ beingvery tiny.)
Deﬁnef (x) on[0, 1] by
f (x) =

1 if 0≤ x <
1
2
−δ
1

x
δ
if
1
2
−δ ≤ x <
1
2
x
δ

1

if
1
2
≤ x <
1
2

1 if
1
2
+δ ≤ x ≤ 1
Thenf iscontinuouson[0, 1]. (Again, just lookat theﬁgures.)
• Theexactareabetweenf andthex-axisisA = 1−
1
2
bh = 1−
1
2
(2δ)(1) = 1−δ. (For δ =
1
10
, wehaveA =
9
10
.)
• With R
1
= 1, theabsoluteerror is |E
1
| = |R
1
−A| = |1−(1−δ)| = δ. (For δ =
1
10
, this absoluteerror is
|E
1
| =
1
10
.)
• WithR
2
=
1
2
, theabsoluteerror is|E
2
| = |R
2
−A| =
¸
¸
1
2
− (1−δ)
¸
¸
=
¸
¸
δ −
1
2
¸
¸
=
1
2
− δ. (For δ =
1
10
, wehave
|E
2
| =
2
5
.)
• Accordingly, R
1
iscloser totheexact areaA thanisR
2
. Indeed, thetinier δ is, themoredramatictheeffect.
• For amonotonicfunction, thisphenomenoncannot occur. Successiveapproximationsfromeither sideget progres-
sivelymoreaccurate.
June 9, 2011 LTSV SSM Second Pass
272 C HA P T E R 5 THE INTEGRAL
x
Right endpt approx, n = 1 Graph of f(x)
0.2 0.4 0.6 0.8 1
1
0.8
0.6
0.4
0.2
0
0.5 1
1
0.5
0
Right endpt approx, n = 2
0.5 1
1
0.5
0
Drawthegraphof apositivecontinuous functiononaninterval suchthat R
2
andL
2
arebothsmaller thanthe
exact areaunder thegraph. Cansuchafunctionbemonotonic?
83. Explaingraphically: The endpoint approximations are less accurate when f

(x) is large.
solution Whenf

islarge, thegraphof f issteeperandhencethereismoregapbetweenf andL
N
orR
N
. Recall that
thetoplinesegmentsof therectanglesinvolvedinanendpoint approximationconstituteapiecewiseconstant function.
If f

islarge, thenf isincreasingmorerapidlyandhenceislesslikeaconstant function.
1 2 4
1
2
3
0 x
y
Smaller f'
3 1 0 0 2 4
1
2
3
0 x
y
Larger f'
3
Provethat for anyfunctionf (x) on[a, b],
R
N
−L
N
=
b −a
N
(f (b) −f (a))
85. Inthisexercise, weprovethat lim
N→∞
R
N
and lim
N→∞
L
N
exist andareequal if f (x) isincreasing[thecase
of f (x) decreasingissimilar]. Weusetheconcept of aleast upper bounddiscussedinAppendixB.
(a) ExplainwithagraphwhyL
N
≤ R
M
for all N, M ≥ 1.
(b) By(a), thesequence{L
N
} isbounded, soit hasaleast upper boundL. Bydeﬁnition, L isthesmallest number such
that L
N
≤ L for all N. Showthat L ≤ R
M
for all M.
(c) Accordingto(b), L
N
≤ L ≤ R
N
for all N. UseEq. (8) toshowthat lim
N→∞
L
N
= L and lim
N→∞
R
N
= L.
solution
(a) Let f (x) bepositiveandincreasing, andlet N andM bepositiveintegers. Fromtheﬁgurebelowat theleft, wesee
that L
N
underestimatestheareaunder thegraphof y = f (x), whilefromtheﬁgurebelowat theright, weseethat R
M
overestimatestheareaunder thegraph. Thus, for all N, M ≥ 1, L
N
≤ R
M
.
x
y
x
y
(b) Becausethesequence{L
N
} is bounded aboveby R
M
for any M, each R
M
is an upper bound for thesequence.
Furthermore, thesequence{L
N
} must havealeast upper bound, call it L. By deﬁnition, theleast upper boundmust be
nogreater thananyother upper bound; consequently, L ≤ R
M
for all M.
(c) SinceL
N
≤ L ≤ R
N
, R
N
−L ≤ R
N
−L
N
, so|R
N
−L| ≤ |R
N
−L
N
|. Fromthis,
lim
N→∞
|R
N
−L| ≤ lim
N→∞
|R
N
−L
N
|.
ByEq. (8),
lim
N→∞
|R
N
−L
N
| = lim
N→∞
1
N
|(b −a)(f (b) −f (a))| = 0,
so lim
N→∞
|R
N
−L| ≤ |R
N
−L
N
| = 0, hence lim
N→∞
R
N
= L.
Similarly, |L
N
−L| = L −L
N
≤ R
N
−L
N
, so
|L
N
−L| ≤ |R
N
−L
N
| =
(b −a)
N
(f (b) −f (a)).
June 9, 2011 LTSV SSM Second Pass
S E C T I ON 5.1 Approximating and Computing Area 273
Thisgivesusthat
lim
N→∞
|L
N
−L| ≤ lim
N→∞
1
N
|(b −a)(f (b) −f (a))| = 0,
so lim
N→∞
L
N
= L.
Thisproves lim
N→∞
L
N
= lim
N→∞
R
N
= L.
UseEq. (8) toshowthat if f (x) ispositiveandmonotonic, thentheareaA under itsgraphover [a, b] satisﬁes
|R
N
−A| ≤
b −a
N
|f (b) −f (a)|
In Exercises 87 and 88, use Eq. (9) to ﬁnd a value of N such that |R
N
−A| < 10
−4
for the given function and interval.
87. f (x) =

x, [1, 4]
solution Let f (x) =

x on[1, 4]. Thenb = 4, a = 1, and
|R
N
−A| ≤
4−1
N
(f (4) −f (1)) =
3
N
(2−1) =
3
N
.
Weneed
3
N
< 10
−4
, whichgivesN > 30,000. Thus|R
30,001
−A| < 10
−4
for f (x) =

x on[1, 4].
f (x) =
_
9−x
2
, [0, 3]
89. Provethat if f (x) ispositiveandmonotonic, thenM
N
liesbetweenR
N
andL
N
andiscloser totheactual
areaunder thegraphthanbothR
N
andL
N
. Hint: Inthecasethat f (x) isincreasing, Figure19showsthatthepartof the
error inR
N
duetotheithrectangleis thesumof theareasA + B + D, andfor M
N
it is |B − E|. Ontheother hand,
A ≥ E.
x
x
i − 1
x
i midpoint
A
F
D
E
B
C
FIGURE 19
solution Supposef (x) ismonotonicincreasingontheinterval [a, b], x =
b −a
N
,
{x
k
}
N
k=0
= {a, a +x, a +2x, . . . , a +(N −1)x, b}
and
_
x

k
_
N−1
k=0
=
_
a +(a +x)
2
,
(a +x) +(a +2x)
2
, . . . ,
(a +(N −1)x) +b
2
_
.
Notethat x
i
< x

i
< x
i+1
impliesf (x
i
) < f (x

i
) < f (x
i+1
) for all 0≤ i < N becausef (x) ismonotoneincreasing.
Then

L
N
=
b −a
N
N−1

k=0
f (x
k
)

<

M
N
=
b −a
N
N−1

k=0
f (x

k
)

<

R
N
=
b −a
N
N

k=1
f (x
k
)

Similarly, if f (x) ismonotonedecreasing,

L
N
=
b −a
N
N−1

k=0
f (x
k
)

>

M
N
=
b −a
N
N−1

k=0
f (x

k
)

>

R
N
=
b −a
N
N

k=1
f (x
k
)

Thus, if f (x) ismonotonic, thenM
N
alwaysliesinbetweenR
N
andL
N
.
Now, as inFigure19, consider thetypical subinterval [x
i−1
, x
i
] andits midpoint x

i
. Welet A, B, C, D, E, andF
be the areas as shown in Figure 19. Note that, by the fact that x

i
is the midpoint of the interval, A = D + E and
F = B +C. LetE
R
representtherightendpointapproximationerror ( = A+B +D), letE
L
representtheleftendpoint
approximationerror ( = C +F +E) andlet E
M
represent themidpoint approximationerror ( = |B −E|).
• If B > E, thenE
M
= B −E. Inthiscase,
E
R
−E
M
= A +B +D −(B −E) = A +D +E > 0,
soE
R
> E
M
, while
E
L
−E
M
= C +F +E −(B −E) = C +(B +C) +E −(B −E) = 2C +2E > 0,
so E
L
> E
M
. Therefore, themidpoint approximationis moreaccuratethaneither theleft or theright endpoint
approximation.
June 9, 2011 LTSV SSM Second Pass
274 C HA P T E R 5 THE INTEGRAL
• If B < E, thenE
M
= E −B. Inthiscase,
E
R
−E
M
= A +B +D −(E −B) = D +E +D −(E −B) = 2D +B > 0,
sothat E
R
> E
M
while
E
L
−E
M
= C +F +E −(E −B) = C +F +B > 0,
so E
L
> E
M
. Therefore, themidpoint approximationis moreaccuratethaneither theright or theleft endpoint
approximation.
• If B = E, themidpoint approximationisexactlyequal tothearea.
Hence, for B < E, B > E, or B = E, themidpoint approximationismoreaccuratethaneither theleft endpoint or the
right endpoint approximation.
5.2 The Definite Integral
Preliminary Questions
1. What is
_
5
3
dx [thefunctionisf (x) = 1]?
solution
_
5
3
dx =
_
5
3
1· dx = 1(5−3) = 2.
2. Let I =
_
7
2
f (x) dx, wheref (x) iscontinuous. Statewhether trueor false:
(a) I istheareabetweenthegraphandthex-axisover [2, 7].
(b) If f (x) ≥ 0, thenI istheareabetweenthegraphandthex-axisover [2, 7].
(c) If f (x) ≤ 0, then−I istheareabetweenthegraphof f (x) andthex-axisover [2, 7].
solution
(a) False.
_
b
a
f (x) dx isthesigned areabetweenthegraphandthex-axis.
(b) True.
(c) True.
3. Explaingraphically:
_
π
0
cosx dx = 0.
solution Becausecos(π − x) = −cosx, the“negative” areabetweenthegraphof y = cosx andthex-axis over
[
π
2
, π] exactlycancelsthe“positive” areabetweenthegraphandthex-axisover [0,
π
2
].
4. Whichisnegative,
_
−5
−1
8dx or
_
−1
−5
8dx?
solution Because−5−(−1) = −4,
_
−5
−1
8dx isnegative.
Exercises
In Exercises 1–10, draw a graph of the signed area represented by the integral and compute it using geometry.
1.
_
3
−3
2x dx
solution Theregionboundedby thegraphof y = 2x andthex-axis over theinterval [−3, 3] consists of tworight
triangles. Onehasarea
1
2
(3)(6) = 9belowtheaxis, andtheother hasarea
1
2
(3)(6) = 9abovetheaxis. Hence,
_
3
−3
2x dx = 9−9= 0.
−3 −2
−2
−4
−6
−1 1 2 3
2
4
6
x
y
June 9, 2011 LTSV SSM Second Pass
S E C T I ON 5.2 The Definite Integral 275
_
3
−2
(2x +4) dx
3.
_
1
−2
(3x +4) dx
solution Theregionboundedby thegraphof y = 3x + 4andthex-axis over theinterval [−2, 1] consists of two
right triangles. Onehasarea
1
2
(
2
3
)(2) =
2
3
belowtheaxis, andtheother hasarea
1
2
(
7
3
)(7) =
49
6
abovetheaxis. Hence,
_
1
−2
(3x +4) dx =
49
6

2
3
=
15
2
.
−2
−2
−1 1
2
4
8
6
x
y
_
1
−2
4dx
5.
_
8
6
(7−x) dx
solution Theregionboundedbythegraphof y = 7− x andthex-axisover theinterval [6, 8] consistsof tworight
triangles. Onetrianglehasarea
1
2
(1)(1) =
1
2
abovetheaxis, andtheother hasarea
1
2
(1)(1) =
1
2
belowtheaxis. Hence,
_
8
6
(7−x) dx =
1
2

1
2
= 0.
−1
8 6 4 2
0.5
−0.5
1
x
y
_
3π/2
π/2
sinx dx
7.
_
5
0
_
25−x
2
dx
solution Theregionboundedbythegraphof y =
_
25−x
2
andthex-axisover theinterval [0, 5] isone-quarter of
acircleof radius5. Hence,
_
5
0
_
25−x
2
dx =
1
4
π(5)
2
=
25π
4
.
5 4 3 2 1
3
4
5
1
2
x
y
_
3
−2
|x| dx
9.
_
2
−2
(2−|x|) dx
solution Theregionboundedbythegraphof y = 2−|x| andthex-axisover theinterval [−2, 2] isatriangleabove
theaxiswithbase4andheight 2. Consequently,
_
2
−2
(2−|x|) dx =
1
2
(2)(4) = 4.
−2 −1 2 1
2
1
x
y
June 9, 2011 LTSV SSM Second Pass
276 C HA P T E R 5 THE INTEGRAL
_
5
−2
(3+x −2|x|) dx
11. Calculate
_
10
0
(8−x) dx intwoways:
(a) Asthelimit lim
N→∞
R
N
(b) Bysketchingtherelevant signedareaandusinggeometry
solution Let f (x) = 8−x over [0, 10]. Consider theintegral
_
10
0
f (x) dx =
_
10
0
(8−x) dx.
(a) Let N beapositiveinteger andset a = 0, b = 10, x = (b −a) /N = 10/N. Also, let x
k
= a + kx = 10k/N,
k = 1, 2, . . . , N betheright endpointsof theN subintervalsof [0, 10]. Then
R
N
= x
N

k=1
f (x
k
) =
10
N
N

k=1
_
8−
10k
N
_
=
10
N

8

N

k=1
1

10
N

N

k=1
k

=
10
N
_
8N −
10
N
_
N
2
2
+
N
2
__
= 30−
50
N
.
Hence lim
N→∞
R
N
= lim
N→∞
_
30−
50
N
_
= 30.
(b) Theregionboundedbythegraphof y = 8−x andthex-axisover theinterval [0, 10] consistsof tworighttriangles.
Onetrianglehasarea
1
2
(8)(8) = 32abovetheaxis, andtheother hasarea
1
2
(2)(2) = 2belowtheaxis. Hence,
_
10
0
(8−x) dx = 32−2= 30.
2
2
4
6
8
4 6 8
10
y
x
Calculate
_
4
−1
(4x −8) dx intwoways: Asthelimit lim
N→∞
R
N
andusinggeometry.
In Exercises 13 and 14, refer to Figure 14.
y = f (x)
6 4 2
y
x
FIGURE 14 Thetwopartsof thegrapharesemicircles.
13. Evaluate: (a)
_
2
0
f (x) dx (b)
_
6
0
f (x) dx
solution Let f (x) begivenbyFigure14.
(a) Thedeﬁniteintegral
_
2
0
f (x) dx isthesignedareaof asemicircleof radius1whichliesbelowthex-axis. Therefore,
_
2
0
f (x) dx = −
1
2
π (1)
2
= −
π
2
.
(b) Thedeﬁniteintegral
_
6
0
f (x) dx is thesigned areaof asemicircleof radius 1 which lies below thex-axis and a
semicircleof radius2whichliesabovethex-axis. Therefore,
_
6
0
f (x) dx =
1
2
π (2)
2

1
2
π (1)
2
=

2
.
Evaluate: (a)
_
4
1
f (x) dx (b)
_
6
1
|f (x)| dx
In Exercises 15 and 16, refer to Figure 15.
1 2 3 4 5
2
1
−1
−2
y = g(t)
t
y
FIGURE 15
June 9, 2011 LTSV SSM Second Pass
S E C T I ON 5.2 The Definite Integral 277
15. Evaluate
_
3
0
g(t ) dt and
_
5
3
g(t ) dt .
solution
• Theregionboundedbythecurvey = g(x) andthex-axisover theinterval [0, 3] iscomprisedof tworighttriangles,
one with area
1
2
below the axis, and one with area 2 above the axis. The deﬁnite integral is therefore equal to
2−
1
2
=
3
2
.
• Theregionboundedbythecurvey = g(x) andthex-axisover theinterval [3, 5] iscomprisedof another tworight
triangles, onewitharea1abovetheaxisandonewitharea1belowtheaxis. Thedeﬁniteintegral isthereforeequal
to0.
Finda, b, andc suchthat
_
a
0
g(t ) dt and
_
c
b
g(t ) dt areaslargeaspossible.
17. DescribethepartitionP andthesetof samplepointsC for theRiemannsumshowninFigure16. Computethevalue
of theRiemannsum.
x
1 3 2.5 3.2 2 0.5 4.5 5
34.25
20
15
8
y
FIGURE 16
solution ThepartitionP isdeﬁnedby
x
0
= 0 < x
1
= 1 < x
2
= 2.5 < x
3
= 3.2 < x
4
= 5
Theset of samplepointsisgivenbyC = {c
1
= 0.5, c
2
= 2, c
3
= 3, c
4
= 4.5}. Finally, thevalueof theRiemannsumis
34.25(1−0) +20(2.5−1) +8(3.2−2.5) +15(5−3.2) = 96.85.
ComputeR(f, P, C) for f (x) = x
2
+x for thepartitionP andtheset of samplepointsC inFigure16.
In Exercises 19–22, calculate the Riemann sum R(f, P, C) for the given function, partition, and choice of sample points.
Also, sketch the graph of f and the rectangles corresponding to R(f, P, C).
19. f (x) = x, P = {1, 1.2, 1.5, 2}, C = {1.1, 1.4, 1.9}
solution Let f (x) = x. With
P = {x
0
= 1, x
1
= 1.2, x
2
= 1.5, x
3
= 2} and C = {c
1
= 1.1, c
2
= 1.4, c
3
= 1.9},
weget
R(f, P, C) = x
1
f (c
1
) +x
2
f (c
2
) +x
3
f (c
3
)
= (1.2−1)(1.1) +(1.5−1.2)(1.4) +(2−1.5)(1.9) = 1.59.
Hereisasketchof thegraphof f andtherectangles.
0.5 1 1.5 2 2.5
0.5
1
2
1.5
x
y
f (x) = 2x +3, P = {−4, −1, 1, 4, 8}, C = {−3, 0, 2, 5}
21. f (x) = x
2
+x, P = {2, 3, 4.5, 5}, C = {2, 3.5, 5}
solution Let f (x) = x
2
+x. With
P = {x
0
= 2, x
1
= 3, x
3
= 4.5, x
4
= 5} and C = {c
1
= 2, c
2
= 3.5, c
3
= 5},
weget
R(f, P, C) = x
1
f (c
1
) +x
2
f (c
2
) +x
3
f (c
3
)
= (3−2)(6) +(4.5−3)(15.75) +(5−4.5)(30) = 44.625.
Hereisasketchof thegraphof f andtherectangles.
June 9, 2011 LTSV SSM Second Pass
278 C HA P T E R 5 THE INTEGRAL
5
10
15
20
25
30
y
x
5 1 4 2 3
f (x) = sinx, P =
_
0,
π
6
,
π
3
,
π
2
_
, C = {0.4, 0.7, 1.2}
In Exercises 23–28, sketch the signed area represented by the integral. Indicate the regions of positive and negative area.
23.
_
5
0
(4x −x
2
) dx
solution Hereisasketchof thesignedarearepresentedbytheintegral
_
5
0
(4x −x
2
) dx.
1 2 3 4
5
−4
−2
2
4
y
x
_
π/4
−π/4
tanx dx
25.
_

π
sinx dx
solution Hereisasketchof thesignedarearepresentedbytheintegral
_

π
sinx dx.
−0.4
−0.8
−1.2
7 5 3 1 6 4 2
0.4
x
y

_

0
sinx dx
27.
_
6
0
(|12−4x| −4) dx
solution Hereisasketchof thesignedarearepresentedbytheintegral
_
6
0
(|12−4x| −4) dx:
+ +

1
−2
2
0
4
6
8
2 3 4 5 6
x
y
_
2
−2
(t
2
−1)(t
2
−4) dt
In Exercises 29–32, determine the sign of the integral without calculating it. Draw a graph if necessary.
29.
_
1
−2
x
4
dx
solution Theintegrand is always positive. Theintegral must thereforebepositive, sincethesigned areahas only
positivepart.
_
1
−2
x
3
dx
31.
_

0
x sinx dx
solution Asyoucanseefromthegraphbelow, theareabelowtheaxisisgreater thantheareaabovetheaxis. Thus,
thedeﬁniteintegral isnegative.
June 9, 2011 LTSV SSM Second Pass
S E C T I ON 5.2 The Definite Integral 279
−0.2
−0.4
−0.6
7 5 3 1 6 4 2
0.2
x
y

+
_

0
sinx
x
dx
In Exercises 33–42, use properties of the integral and the formulas in the summary to calculate the integrals.
33.
_
4
0
(6t −3) dt
solution
_
4
0
(6t −3) dt = 6
_
4
0
t dt −3
_
4
0
1dt = 6·
1
2
(4)
2
−3(4−0) = 36.
_
2
−3
(4x +7) dx
35.
_
9
0
x
2
dx
solution Byformula(5),
_
9
0
x
2
dx =
1
3
(9)
3
= 243.
_
5
2
x
2
dx
37.
_
1
0
(u
2
−2u) du
solution
_
1
0
(u
2
−2u) du =
_
1
0
u
2
du −2
_
1
0
u du =
1
3
(1)
3
−2
_
1
2
_
(1)
2
=
1
3
−1= −
2
3
.
_
1/2
0
(12y
2
+6y) dy
39.
_
1
−3
(7t
2
+t +1) dt
solution First, write
_
1
−3
(7t
2
+t +1) dt =
_
0
−3
(7t
2
+t +1) dt +
_
1
0
(7t
2
+t +1) dt
= −
_
−3
0
(7t
2
+t +1) dt +
_
1
0
(7t
2
+t +1) dt
Then,
_
1
−3
(7t
2
+t +1) dt = −
_

1
3
(−3)
3
+
1
2
(−3)
2
−3
_
+
_

1
3
1
3
+
1
2
1
2
+1
_
= −
_
−63+
9
2
−3
_
+
_
7
3
+
1
2
+1
_
=
196
3
.
_
3
−3
(9x −4x
2
) dx
41.
_
1
−a
(x
2
+x) dx
solution First,
_
b
0
(x
2
+x) dx =
_
b
0
x
2
dx +
_
b
0
x dx =
1
3
b
3
+
1
2
b
2
. Therefore
_
1
−a
(x
2
+x) dx =
_
0
−a
(x
2
+x) dx +
_
1
0
(x
2
+x) dx =
_
1
0
(x
2
+x) dx −
_
−a
0
(x
2
+x) dx
=
_
1
3
· 1
3
+
1
2
· 1
2
_

_
1
3
(−a)
3
+
1
2
(−a)
2
_
=
1
3
a
3

1
2
a
2
+
5
6
.
_
a
2
a
x
2
dx
In Exercises 43–47, calculate the integral, assuming that
_
5
0
f (x) dx = 5,
_
5
0
g(x) dx = 12
43.
_
5
0
(f (x) +g(x)) dx
solution
_
5
0
(f (x) +g(x)) dx =
_
5
0
f (x) dx +
_
5
0
g(x) dx = 5+12= 17.
June 9, 2011 LTSV SSM Second Pass
280 C HA P T E R 5 THE INTEGRAL
_
5
0
_
2f (x) −
1
3
g(x)
_
dx
45.
_
0
5
g(x) dx
solution
_
0
5
g(x) dx = −
_
5
0
g(x) dx = −12.
_
5
0
(f (x) −x) dx
47. Isit possibletocalculate
_
5
0
g(x)f (x) dx fromtheinformationgiven?
solution It isnot possibletocalculate
_
5
0
g(x)f (x) dx fromtheinformationgiven.
Provebycomputingthelimit of right-endpoint approximations:
_
b
0
x
3
dx =
b
4
4
In Exercises 49–54, evaluate the integral using the formulas in the summary and Eq. (9).
49.
_
3
0
x
3
dx
solution ByEq. (9),
_
3
0
x
3
dx =
3
4
4
=
81
4
.
_
3
1
x
3
dx
51.
_
3
0
(x −x
3
) dx
solution
_
3
0
(x −x
3
) dx =
_
3
0
x dx −
_
3
0
x
3
dx =
1
2
3
2

1
4
3
4
= −
63
4
.
_
1
0
(2x
3
−x +4) dx
53.
_
1
0
(12x
3
+24x
2
−8x) dx
solution
_
1
0
(12x
3
+24x
2
−8x) dx = 12
_
1
0
x
3
dx +24
_
1
0
x
2
−8
_
1
0
x dx
= 12·
1
4
1
4
+24·
1
3
1
3
−8·
1
2
1
2
= 3+8−4= 7
_
2
−2
(2x
3
−3x
2
) dx
In Exercises 55–58, calculate the integral, assuming that
_
1
0
f (x) dx = 1,
_
2
0
f (x) dx = 4,
_
4
1
f (x) dx = 7
55.
_
4
0
f (x) dx
solution
_
4
0
f (x) dx =
_
1
0
f (x) dx +
_
4
1
f (x) dx = 1+7= 8.
_
2
1
f (x) dx
57.
_
1
4
f (x) dx
solution
_
1
4
f (x) dx = −
_
4
1
f (x) dx = −7.
_
4
2
f (x) dx
In Exercises 59–62, express each integral as a single integral.
59.
_
3
0
f (x) dx +
_
7
3
f (x) dx
solution
_
3
0
f (x) dx +
_
7
3
f (x) dx =
_
7
0
f (x) dx.
_
9
2
f (x) dx −
_
9
4
f (x) dx
61.
_
9
2
f (x) dx −
_
5
2
f (x) dx
solution
_
9
2
f (x) dx −
_
5
2
f (x) dx =
_
_
5
2
f (x) dx +
_
9
5
f (x) dx
_

_
5
2
f (x) dx =
_
9
5
f (x) dx.
_
3
7
f (x) dx +
_
9
3
f (x) dx
In Exercises 63–66, calculate the integral, assuming that f is integrable and
_
b
1
f (x) dx = 1−b
−1
for all b > 0.
63.
_
5
1
f (x) dx
solution
_
5
1
f (x) dx = 1−5
−1
=
4
5
.
June 9, 2011 LTSV SSM Second Pass
S E C T I ON 5.2 The Definite Integral 281
_
5
3
f (x) dx
65.
_
6
1
(3f (x) −4) dx
solution
_
6
1
(3f (x) −4) dx = 3
_
6
1
f (x) dx −4
_
6
1
1dx = 3(1−6
−1
) −4(6−1) = −
35
2
.
_
1
1/2
f (x) dx
67. Explainthedifferenceingraphical interpretationbetween
_
b
a
f (x) dx and
_
b
a
|f (x)| dx.
solution When f (x) takes on both positiveand negativevalues on [a, b],
_
b
a
f (x) dx represents thesigned area
betweenf (x) andthex-axis, whereas
_
b
a
|f (x)| dx represents thetotal (unsigned) areabetweenf (x) andthex-axis.
Anynegativelysignedareasthatwerepartof
_
b
a
f (x) dx areregardedaspositiveareasin
_
b
a
|f (x)| dx. Hereisagraphical
exampleof thisphenomenon.
−20
2 4 −4 −2
10
−30
−10
x
Graph of f (x)
2 4 −4 −2
10
20
30
x
Graph of |f (x)|
Usethegraphical interpretationof thedeﬁniteintegral toexplaintheinequality
¸
¸
¸
¸
¸
_
b
a
f (x) dx
¸
¸
¸
¸
¸

_
b
a
|f (x)| dx
wheref (x) iscontinuous. Explainalsowhy equality holdsif andonly if either f (x) ≥ 0for all x or f (x) ≤ 0for
all x.
69. Let f (x) = x. Findaninterval [a, b] suchthat
¸
¸
¸
¸
¸
_
b
a
f (x) dx
¸
¸
¸
¸
¸
=
1
2
and
_
b
a
|f (x)| dx =
3
2
solution If a > 0, thenf (x) ≥ 0for all x ∈ [a, b], so
¸
¸
¸
¸
¸
_
b
a
f (x) dx
¸
¸
¸
¸
¸
=
_
b
a
|f (x)| dx
bythepreviousexercise. Weﬁndasimilar result if b < 0. Thus, wemust havea < 0andb > 0. Now,
_
b
a
|f (x)| dx =
1
2
a
2
+
1
2
b
2
.
Because
_
b
a
f (x) dx =
1
2
b
2

1
2
a
2
,
then
¸
¸
¸
¸
¸
_
b
a
f (x) dx
¸
¸
¸
¸
¸
=
1
2
|b
2
−a
2
|.
If b
2
> a
2
, then
1
2
a
2
+
1
2
b
2
=
3
2
and
1
2
(b
2
−a
2
) =
1
2
yielda = −1andb =

2. Ontheother hand, if b
2
< a
2
, then
1
2
a
2
+
1
2
b
2
=
3
2
and
1
2
(a
2
−b
2
) =
1
2
yielda = −

2andb = 1.
EvaluateI =
_

0
sin
2
x dx andJ =
_

0
cos
2
x dx as follows. First showwithagraphthat I = J. Then
provethat I +J = 2π.
In Exercises 71–74, calculate the integral.
71.
_
6
0
|3−x| dx
solution Over theinterval, theregionbetweenthecurveandtheinterval [0, 6] consistsof twotrianglesabovethex
axis, eachof whichhasheight 3andwidth3, andsoarea
9
2
. Thetotal area, hencethedeﬁniteintegral, is9.
June 9, 2011 LTSV SSM Second Pass
282 C HA P T E R 5 THE INTEGRAL
6 5 4 3 2 1
1
2
3
x
y
Alternately,
_
6
0
|3−x| dx =
_
3
0
(3−x) dx +
_
6
3
(x −3) dx
= 3
_
3
0
dx −
_
3
0
x dx +
_
_
6
0
x dx −
_
3
0
x dx
_
−3
_
6
3
dx
= 9−
1
2
3
2
+
1
2
6
2

1
2
3
2
−9= 9.
_
3
1
|2x −4| dx
73.
_
1
−1
|x
3
| dx
solution
|x
3
| =
_
x
3
x ≥ 0
−x
3
x < 0.
Therefore,
_
1
−1
|x
3
| dx =
_
0
−1
−x
3
dx +
_
1
0
x
3
dx =
_
−1
0
x
3
dx +
_
1
0
x
3
dx =
1
4
(−1)
4
+
1
4
(1)
4
=
1
2
.
_
2
0
|x
2
−1| dx
75. UsetheComparisonTheoremtoshowthat
_
1
0
x
5
dx ≤
_
1
0
x
4
dx,
_
2
1
x
4
dx ≤
_
2
1
x
5
dx
solution Ontheinterval [0, 1], x
5
≤ x
4
, so, byTheorem5,
_
1
0
x
5
dx ≤
_
1
0
x
4
dx.
Ontheother hand, x
4
≤ x
5
for x ∈ [1, 2], so, bythesameTheorem,
_
2
1
x