Upload
vocong
View
238
Download
0
Embed Size (px)
Citation preview
Answers to Exercises
I am much indebted to Anthony Petrello for some of the answers to the exercises.
I, §2, p. 13
1. - 3 < x < 3 2. - 1 ~ x ~ 0 3. - j3 ~ x ~ - 1 or 1 ~ x ~ j3 4. x < 3 or x > 7 5. - 1 < x < 2 6. x < -lor x > 1 7. - 5 < x < 5 8. - 1 ~ x ~ 0 9. x ~ 1 or x = 0 10. x ~ - 10 or x = 5
11. x ~ -10 or x = 5 12. x ~ 1 or x = -! 13. x < -4 14. -5<x< -3 15. -3<x< -2 and -2<x< -1 16. -2<x<2 17. -2 < x < 8 18. 2 < x < 4 19. -4 < x < 10 20. x < -4 and x> 10 21. x < -10 and x > 4
I, §3, p. 17
1. -~ 2. (2x ~ 1) 3. 0, 2, 108 4. 2z - Z2, 2w - w2
5. x # j2 or -j2. f(5) = b 6. All x. f(27) = 3 7. (a) 1 (b) 1 (c) -1 (d) -1 8. (a) 1 (b) 4 (c) 0 (d) 0 9. (a) -2 (b) -6 (c) x2+4x-2 10. x~O, 2
II. (a) odd (b) even (c) odd (d) odd
I, §4, p. 20
1. 8 and 9 2.! and -1 3. /6 and 2 4. ~ and 21/ 3 5. -h and! 6. 9 and 8 7. -! and -1 8.! and! 9. 1 and -! 10. -5:2 and!
A2 ANSWERS TO EXERCISES
II. Yes. Suppose a is negative, so write a = -b where b is positive. Let c be a positive number such that c" = b. Then (-c)" = a because (-1)" = -1 since n is odd.
II, §1, p. 24
3. x negative, y positive 4. x negative, y negative
II, §3, p. 33
5. y = -~x + i 6. y = -ix + S 7. x = J2 9 9j3
8. y = --;:;- x + 4 - --;:;- 9. y = 4x - 3 10. y = -2x + 2 ",3+3 ",3+3
II. y= -!x+3+ J'!- 12. y=j3x+S+j3 19. -! 20.-8
21. 2 + j2 22.!<3 + j3) 23. y = (x - n>(J22_ n) + 1
24. y=(X-j2)C~-~)+2 25. y= -(x+ l)(j23+ 1)+2
26. y = (x + 1)(3 + j2) + j2 29. (a) x = -4, y = -7 (b) x = ~, y = i (c) x = -!, y = 1 (d) x = - 6, Y = - 5
II, §4, p. 35
1.J97 2.j2 3.Js2 4.Ji3 5.!j5 6.(4,-3) 7. SandS 8.(-2,5) 9. Sand 7
II, §8, p. 51
5. (x - 2)2 + (y + 1)2 = 25 6. x 2 + (y - 1)2 = 9 7. (x + \)2 + y2 = 3 8. y + ¥ = 2(x + W 9. y - 1 = (x + 2)2 10. y + 4 = (x - \)2
II. (x + 1)2 + (y - 2)2 = 2 12. (x - 2)2 + (y - \)2 = 2 13. x + ¥- = 2(y + !)2 14. x-I = (y + 2)2
III, §1, p. 61
1. 4 2. -2 3. 2 4. i S. -! 6. 0 7. 4 8. 6 9. 3 10. 12 II. 2 12. 3 13. a
ANSWERS TO EXERCISES
III, §2, p. 70
Tangent line at x = 2 Slope at x = 2
1. 2x y = 4x - 3 4 2. 3x2 y = 12x - 16 12 3. 6x2 y = 24x - 32 24 4.6x y = 12x - 12 12 5.2x y = 4x - 9 4 6. 4x + I Y = 9x - 8 9 7. 4x - 3 Y = 5x - 8 5
3x2 8. 2+2 y = 8x - 8 8
9. Y = -~x + ~ 1
- (x + 1)2 -9
2 Y = -~x +.!j 2 10. -
(x + 1)2 -9
III, §3, p. 75
2 I 1 I. 4x + 3 2. - 2 3. 2 4. 2x + 1 5. - 2 6. 9x 2
(2x + 1) (x + 1) (2x - 1)
3x2
10. 2 + I 11. _2/X2 12. -3/x2
13. - 2/(2x - W 14. - 3/(3x + 1)2 15. -1/(x + 5)2 16. -I/(x _ 2)2 17. -2x- 3 18. -2(x + 1)-3
III, §4, p. 78
1. X4 + 4x3h + 6x 2h2 + 4xh3 + h4 2. 4x3 3. (a) jx- I /3 (b) _~X-5/ 2 (c) ~XI/6 4. Y = 9x - 8 5. Y = tx + t slope t
-3 7 -3 1 -fi 1 6. Y = Y x + 32' slope y 7. Y = 2-fi x + 2' slope 2-fi
8. (a) !5- 3/4 (b) _!r 5/4 (c) )2(\0)2 - 1) (d) n7n - 1
III, §5, p. 89
l. (a) ~X-2/3 (b) £X- 1/4 (c) x (d) £x 2
2. (a) 55x 'O (b) -8x- 3 (c) ~X3 - 15x2 + 2x 3. (a) -iX- 7/4 (b) 3 - 6x 2 (c) 20x4 - 21x2 + 2 4. (a) 21x2 + 8x (b) ~X-l/3 + 20x 3 - 3x2 + 3 5. (a) -25x- 2 + 6X- I /2 (b) 6x2 + 35x6 (c) 16x3 - 21x2 + 1 6. (a) ~x - 16x 7 (b) 12x3 - 4x + 1 (c) 7nx6 - 40x4 + 1 7. (x 3 + x) + (3x 2 + I )(x - I) 8. (2x 2 - I )4x3 + 4x(x4 + I) 9. (x + 1)(2x + ¥Xl /l) + (Xl + 5XJ/l)
10. (2x - 5)(l2x3 + 5) + 2(3x4 + 5x + 2)
A3
A4 ANSWERS TO EXERCISES
(_2X2 + 2) 14. 2 2
(X + 3x + 1)
(t + 1)(t - 1)(2t + 2) - (t2 + 2t - 1)2t 15. (t2 _ 1)2
(t2 + t - 1)( - 5/4)t- 9/4 - t- 5/4(2t + 1) 16. 2 2
(t + t - 1)
17. i9, y = i9t +;@ 18. !, y = !t
III, §5, Supplementary Exercises, p. 89
1. 9x2 - 4 3. 2x + 1 5. ~.t3 / 2 - ~X-7/2 7. x 2 - 1 + (x + 5X2x) 9. (~XI /2 + 2x)(x4 - 99) + (X 3/2 + x2)(4x3 )
11. (4X{:2 + 4x + 8) + (2x 2 + 1)( ~32 + 4)
13. (x + 2Xx + 3) + (x + 1)(x + 3) + (x + lXx + 2) 15. 3x2(X 2 + 1)(x + 1) + x 3(2x)(x + 1) + (X3XX 2 + 1)
-2 5(3x2 + 4x) -2(x + 1) + 2x 17. 2 19. 3 2 2 21. 2
(2x + 3) (x + 2x ) (x + 1)
(x + lXx - 1)3(!x- I/2) - 3XI /2 [(X - 1) + (x + 1)]
23. (x + 1)2(x _ 1)2
(x 2 + 1)(x + 7)(5x4) - (x 5 + 1)«x2 + 1) + (2x)(x + 7») 25. (x2 + 1)2(x + 7)2
(1 - x 2)(3x2) - x 3( - 2x) (x2 + 1)(2x - 1) - (x2 - x)(2x) 27. (l _ X2)2 29. (x2 + 1)2
(x2 + X - 4)(2) - (2x + 1)(2x + 1) 31. ( 2 4)2 X +x-
(x 2 + 2)(4 - 3x2) - (4x - x 3 )(2x) -5x - (1 - 5x) n 2 22 li (x + )
(x + 1)(x - 2X2x) - x2«x - 2) + (x + 1») 37. 2 2
(x + 1) (x - 2)
(4x 3 - x 5 + 1)(12x 3 + iX' /4) - (3x4 + x5/4)(12x2 - 5x4) 39. (4x3 _ x5 + 1)2
41. (y - 18) = H(x - 16) 43. (y + 12) = 19x 45. (y - to) = 14(x - I)
4 -12 4 -4 47. Y - 9 = 81 (x - 2) 49. y - 3 = 9 (x - 2)
51. Point of tangency: (3, - 3). Both curves intersect here and have slope -1. 53. Both curves have the point (1, 3) in common and have slope 6 at this point. 55. Tangent line (y - 7) = 16x at (0,7); tangent line (y - 19) = 16(x - 1)
at (I, 19); tangent line (y + 13) = 16(x + I) at (-I, -13).
ANSWERS TO EXERCISES
III, §6, p. 99
1. 8(x+ 1)7 2. !(2x-5)-1 /2· 2 3.3(sinx)2cos x 4.5(lOgX)4G)
1 1 5. (cos 2x)2 6. -2--1 (2x) 7. eCOSX( -sin x) 8. . (r + cos x)
x + eX + sm x
9. cos log x + ~ ~ - 2 10. . 2 [ IJ(1 1) sin 2x - (x + l)(cos 2x)2 X X x (sm 2x)
1 11. 3(2x2 + W(4x) 12. -[sin(sin 5x)](cos 5x)5 13. -2- (-sin 2x)2
cos x
14. [cos(2x + W](2(2x + 5»)(2). 15. [cos(cos(x + 1»)]( -sin(x + 1»)
1 3 1 2 16. (cos eX)eX 17. - (3x _ 1)8 [4(3x - \) ]·3 18. - (4X)6' 3(4x) ·4
1 1 19. - (. 2)4 2(sin 2x)(cos 2x)· 2 20. - 2 4 2(cos 2x)( -sin 2x)2
sm x (cos x)
1 21. (. 3 2 (cos 3x)· 3 22. -sin2 x + cos2 x 23. (x2 + l)ex + 2xex
sm x)
24. (x3 + 2x)(cos 3x)· 3 + (3x 2 + 2) sin 3x
1 2ex cos 2x - (sin 2x)eX 25. - ( . )2 (cos X - sin x) 26.
sm x + cos x e2x
(x 2 + 3)/x - (log x)(2x) cos 2x - (x + \)( -sin 2x)· 2 27. 2 2 28. 2
(x + 3) cos 2x
29. (2x - 3)(eX + \) + 2(eX + x) 30. (x3 - 1)(e3x . 3 + 5) + 3x2(e3x + 5x)
(x - 1)3x2 - (x 3 + \) (2x + 3)2x - (x2 - 1)2 31. (x _ 1)2 32. (2x + W 33. 2(X4/3 - eX) + (~XI /3 - eX)(2x + 1)
34. (sin 3x)ix- 3/4 + 3(cos 3X)(X I /4 - 1) 35. [cos(x2 + 5x)](2x + 5)
3x' +8 -I 1 3 36. e (6x) 37. [Iog(X4 + 0]2' X4 + 1 ·4x
- 1 1 I _ 1/2 2ex - 2xex 38. [Iog(X1 /2 + 2X)]2 (x1 /2 + 2x) (IX + 2) 39. e2x
2x . 4 40. -I - -6' 65 . +x
III, §6, Supplementary Exercises, p. 100
I. 2(2x + 1)2 3. 7(5x + 3)65 5. 3(2x2 + X - W(4x + 1) 7. !(3x + 1) - 1/2(3) 9. -2(x2 + X - 0-3(2x + 1) 11. -i(x + 5) - 8/3
13. (x - 1)3(x - W + (x - WIS. 4(x3 + x2 - 2x - 1)3(3x2 + 2x - 2)
17. (x - 1)1 /2(i)(x + 0- 1/4 - (x + 1)3/4<t)(x -0- 1/2
x-I
AS
A6 ANSWERS TO EXERCISES
(3x + 2)9(~)(2x2 + X - 1)3/2(4x + 1) - (2X2 + X - 1)5/2(9)(3x + 2)8(3) 19. (3x+2)18
21. i(2x + 1) - 1/2(2) 23. i(x2 + X + 5) - 1/2(2x + 1) 25. 3x2cos(X3 + 1) 27. (exJ +1)(3x2) 29. (cos(cosx»)(-sin x) 31. (esin(x J + 1l)(3x2 cos(x3 + 1)) 33. [cos«x + 1)(x2 + 2»)][(x + 1)(2x) + (x2 + 2)] 35. (e(x+ I)(x - 3»«X + 1) + (x - 3» 37. 2 cos(2x + 5)
39. _ 2 _ 41. (COS x - 5 )(2X + 4) - (x - 5)2) 2x + 1 2x + 4 (2x + 4)2
43. (e2x2+3X+I)(4x + 3) 45. 2x ~ 1 [cos(log2x + 1)]2
47. - (6x - 2) sin(3x2 - 2x + 1) 49. 80(2x + 1)79(2) 51. 49(log x )48(X - I) 53. 5( e2x + 1 - x )4(2e2x + 1 - 1)
55. !(310g(x2 + 1) - X3) - 1/2(_3_ (2x) - 3x 2) 2 x 2 + 1
2(cos 3x)(cos 2x) - 3(sin 2x)( -sin 3x) 57. 2
(cos 3x)
(sin x3)(1 /2x2)4x - (log 2X2)(COS x 3)3x2 59. . 3 2
(SIO X )
(cos 2x)(4x3) - 2(X4 + 4)( -sin 2x) 61. 2
(cos 2x)
(cos x3)(4)(2x2 + 1)3(4x) + (2x2 + 1)4(sin x3)(3x2) 63. 2 3
cos x
65. _3e - 3x 67. (e- 4x2 +X)(_8x + 1)
e- X[2x/(x2 + 2)] - [log(x2 + 2)](e - X)( -1) 69. 2 e x
III, §7, p. 103
I. 18x 2. 5(x2 + 1)4·2 + 20(x2 + 1)34x2 3. 0 4. 7! 5. 0 6. 6 7. -cos x 8. cos x 9. -sin x 10. -cos x 11. sin x 12. cos x
In Problems 7 through 12, there is a pattern. Note that the derivatives of sin x are :
f(x) = sin x;
f(l)(x) = cos x;
j<2)(X) = -sin x;
j<3)(X) = -cos x;
j<4)(X) = sin x.
Then the derivatives repeat. Thus every fourth derivative is the same. Hence to find the n-th derivative, we just divide n by 4, and if r is the remainder, so
ANSWERS TO EXERCISES
n = 4q + r, then
13. (a) 5! (b) 7! (c) 13! 14. (a) k! (b) k! (c) 0 (d) 0
III, §8, p. 106
3 - x y - 3x 2 6x 2 1 - 2y I. -(2x + y)/x 2. -- 3. -2- 4. 2 S. ---=----
Y + 1 3y - x 3y + 1 2x + 2y - 1
2 2 1 + 4xy x(i - 1) 6. - y /x 7. - 2(y + x2 ) 8. y(l _ x2 ) 9. (y - 3) = 3(x + 1)
10. y + 1 = 4(x - 3) II. (y - 2) = i(x - 6) 12. y + 2 = -£(x - I) 13. (y + 4) = £(x - 3) 14. y - 2 = -t<x + 4) IS. y - 3 = ~x - 2)
III, §9, p. 114
I. (a) 1/6 (b) 0 (c) Impossible 2. 0 3. 320 ft/sec2 4. 0 S. 240 m3/sec
3 1tdc.L./ 6. 361t em /sec 7. hr, -, - 8. 16 umts sec 2 h
9. (a) 1 ftjsec (b) 1 ft/sec 10. (a) i ft/sec (b) 2 ft/sec II. (t, H 12. The picture is as follows.
We are given dx/dt = - I and dy/dt = 2. The area is A = txy, so
- = - x - + y - = -[2x - y]. dA 1 [dY dX] 1
dt 2 dt dt 2
A7
We are given that at some time, x = 8. Since the speed is uniform toward the origin, after 2 min we find x = 8 - 2 = 6. Also after 2 min we find y = 6 + 4 = 10. Hence after 2 min we get
dA dt = t[12 - 10] = 1 cm2/min.
13. 90 cm2/sec 14. -815 ft/sec IS. 3/200 ft/min 16. 5/41t ft/min 17. t = t ace = 4
A8 ANSWERS TO EXERCISES
18. Both x and yare functions of time t. Differentiating each side of
y = x 2 - 6x
with respect to t, we find
dy dx dx -=2x --6-. dt dt dt
When dy/dt = 4 dx/dt this yields
dx dx dx 4-=2x--6-.
dt dt dt
Canceling dx/dt yields 4 = 2x - 6, so x = 5, and y = - 5. 19. 4/75n ft/min. Draw the picture
When h = 5, then the distance from the top of the water to the top of the hemisphere is also 5, so by Pythagoras,
52 + r2 = 102.
You can then solve for r. Use that dV/dt = 4 to find dh/dt. 20. 1(502.7 + 602.3)(502.3.52 + 602 . 1.52) -1/2
21. 1/ 12n ft/min
h
----- --r
The assumption about the diameter implies that r = 3h/2 so V = ~nh3. Then
dV 9 2 dh - = -nh - = 3. dt 4 dt
When h = 4 this gives the answer. 22. -3/400 cm/min, -6n/5 cm2/min 23. 1/32n m/min 24. lOOn ft 3/sec
IV, §1, p. 131
1. fi /2 2. fi/2 3. fi/2 4. 1 5. -fi/2 6. -1 7. fi /2 8. -fi/2
9. 1 10. fi II. 1 12. -1 13. -1 14. -fi/2 15. -1/2 16. fi/2
17. -fi/2 18. 1
ANSWERS TO EXERCISES A9
IV, §2, p. 135
2.
I I I
-::---'I:--_---1I--__ t-__ +-__ -,----__ +-_~--'I~x-axis 31t I -7t 1t 1t I 1t 31t1 -2: 2: 2: 2:
I I
(\i (\i 3. y-axis
----,:::+---~--_+--4II...--___j-- x-axis - 1t 1
I
4. (a) (b)
Graph of y = sin 2x Graph of y = sin 3x
AlO ANSWERS TO EXERCISES
5. (a) Graph of y = sin tx
-It
(b) Graph of y = sin !x
21t 31t
6. (a) Graph of y = sin !tx
1 3
(c) y = sin 21tx
7. (a) y = Isin xl
2 2 2
ANSWERS TO EXERCISES All
IV, §3, p. 140
ji;; -ji;; 1. 4 (y' 3 + l) 2. -4- (y' 3 - I)
3. (a) Jj (J3 - l) (b) Jj (J3 + I) (c) Jj (J3 + I) (d) Jj (J3 - l)
(e) Jj (J3 - l) (f) -f (J3 + l) (g)! (h) -J3/2
5. sin 3x == 3 sin x - 4 sin3 x , cos 3x = 4 cos3 X - 3 cos x
IV, §4, p. 145
I. _csc l X 2. 3 cos 3x 3. -5 sin 5x 4. (8x + l) COS(4X2 + x) 5. (3x 2) sec2(x3 - 5) 6. (4x 3 - 3x2 ) sec2(x· - x 3 )
7. cos x secl(sin x) 8. secl x cos(tan x) 9. _sec l x sin(tan x) 10.-1
11. 0 12. J3/2 13. - J2 14. 2 15. - 2J3
16. (a) y = 1 (b) (y - f) = ~ 1 (x - ~) (c) y = 1
(e) y == 1
(g) (y-I)= -2(X-~)
(i) (y - !) == ~ (x - j) (k) y = 1
17. (a) -1.6 (b) S (c) J3/30
18. y-axis
(d) (y + I) = 6( x - ~)
(f) (y - J2) = -J2(x -~) 1t
(h) y - 1 = x - -2
-1tJ3 (j) (y - 1) = -- (x - 1)
6
(I) (y -~) = ~1t (x -n
----,--J---+--.....J.....--II--.....L.-x·axis
Al2 ANSWERS TO EXERCISES
One revolution every two minutes is half a revolution per minute. Hence dO/dt = n (in radians per minute). But
y = 25 sin 0
so dy d sin 0 dO - = 25 -- - = 25(cos O)n. dt dO dt
When the height of a point on the wheel is 42.5 then y = 42.5 - 30 = 12.5. Therefore sin 0 = 12.5/25 = ! and 0 = n/6. Hence
dy I = 25(COS ~)n = ¥ fin ft/min. dt 8=n/6 6
19. (a) 180 ft/sec (b) 360 ft/sec (c) 2250 ft/sec (d) 9000/91 ft/sec (e) 1530 ft/sec 20. 25 rad/hr
21.
We are given dO/dt = 4n (two revolutions = 4n radians). Then
tan 0 = - y-1000
so y = 1000 tan O.
Using the chain rule,
dy dO - = 1000(1 + tan2 0) -. dt dt
(a) The point on the wall nearest to the light is when 0 = O. Then tan 0 = 0, so
dy I . -d = 1000· 4n = 4,OOOn ft/mm. t 8=0
(b) When y = 500 then tan 0 = !, so we substitute! for tan 0 and get
- = l000(i)4n = 5000n ft/mm. dy I . dt y= 500
22. 8000n/27 ft/sec
ANSWERS TO EXERCISES
23. Let s be the length of the shadow.
s
-30-
We are given dO/dt = 1t/ 1O. We also have
30 - s s tan 0 = - - = 1 - - .
30 30
Differentiating with respect to t and using the chain rule gives:
2 dO 1 ds (1 + tan 0) - = - - - .
dt 30 dt
If 0 = 1t/6 then tan(1t/6) = 1/,}i Substituting yields:
We can solve for ds/dt, namely
ds I (4) 1t dt 9=n/6 = -30:3 10 = -41t ft/hr.
24. 54/51t deg/min 25. We are given dx/dt = 3. Find dO/dt when x = 15.
r y
1
We have
-x-x
cos 0 = 30'
A13
A14 ANSWERS TO EXERCISES
Hence by the chain rule,
dO I dx - sin 0 dt = 30 dt '
When x = IS we have cos 0 = 1/2 and so 0 = Tt/3. Hence sin 0 = j3/2. Then
dOl I 1 dt x=15 = - j3/2 303
-I = - - rad/sec.
5j3
26. 9/2n deg/sec
27.
1 200
1
We are given dx/dt = 20. We want to find dO/dt when z = 400. We have
200 tan 0 = - ,
x
so differentiating with respect to t and using the chain rule,
(1+tan2 0) - =200 - - . dO (-I) dx dt x 2 dt
When z = 400, then sin 0 = 200/400 = 1/2. Hence 0 = n/6 and therefore
tan 0 = 1/ j3. Also x = 200j3. This gives
Hence finally
( 1 + D ~~ Iz = 400 = 200 · 2;/ 3 20
-I
30
dOl = - ~ ~ = - ~ ft/sec. dt z=400 30 4 40
28. -1/25 rad/sec
ANSWERS TO EXERCISES A15
IV, §6, p. 157
3. (a) (fl, n/4) (b) (fl, 571:/4) (c) (6, 71:/3) (d) (1, n) 4. (a) (y - 1)2 + x2 = 1 (b) (x _ ~)2 + y2 = £
5. (a) (Y-~Y +x2=GY (b) (x-n2 +y2=GY (c) x2 + (y - a)2 = a2 (d) (x _ a)2 + y2 = a2
6. r2 = cos (J. This is equivalent with r = Jcos (J. Only values of (J such that cos (J ~ 0 will give a contribution to r. Also, since cos( - (J) = cos (J the curve is symmetric with respect to the x-axis. We make a table.
(J r
o to 71:/ 2 dec. 1 to 0 -n/ to 0 inc. 0 to 1
In these intervals, we have 0 ~ cos (J ~ 1, and hence
Jcos (J ~ cos (J, (watch out!)
with equality only at the end points. Since r = cos (J is the equation of a circle (similarly to r = sin (J, see Problem 5), the graph looks like this.
Graph of r1 = cos 8.
8. (a) r = sin2 (J. The right-hand side is always ~ 0 so there is a value of r for each value of (J. Since
-1 ~ sin (J ~ 1,
Al6 ANSWERS TO EXERCISES
it follows that sin2 e ~ Isin el. Also the regions of increase and decrease are over intervals of length n/2. You should make a table of these, and then see that the graph looks like this.
The ovals are thinner than circles, contrary to Problem 6, where they were more expanded than circles.
9. r = 4 sin2 e. Note that the right-hand side is always ~ 0, and so there is a value of r for each value of e. Also
sine -e) = -sin 8 and
so the graph is symmetric with respect to the x-axis. We make a table.
e r
o to n/2 inc. 0 to 4 n/2 to n dec. 4 to O.
Also observe that sin2 8 ~ Isin 01 because Isin 81 ~ 1. Hence the graph is something like this:
10. x2 + y2 = 25 (circle) 11. x2 + y2 = 16 (circle)
ANSWERS TO EXERCISES Ai7
12. (a) rcos8 = 1 is equivalent with x = 1, which is a vertical line! 13. r = 3/cos O. This is defined for cos 8 # O. In this case, the equation is equiva
lent with r cos 0 = 3, and x = r cos 0, so the equation in rectangular coordinates is x = 3, which is a vertical line.
16. (x 2 + x + i)2 = x 2 + i 17. (x 2 + y2 + 2y)2 = x 2 + y2 27. i = 2x + 1 2
28. r = O. We make a table: 2 - cos
0 cos 0 2 - cos 0 r
inc. 0 to n/2 dec. I to 0 inc. I to 2 dec. 2 to 1 inc. n/2 to n dec. 0 to -1 inc. 2 to 3 dec. 1 to 2/3 inc. n to 3n/2 inc. -1 to 0 dec. 3 to 2 inc. 2/3 to 1 inc. 5n/2 to 2n inc. 0 to 1 dec. 2 to 1 inc. 1 to 2
-2 -I 2
-2
One can see that this equation is an ellipse by converting to (x, y)-coordinates. The equation is equivalent with
r(2 - cos 8) = 2, that is 2) x 2 + y2 - X = 2.
By algebra, this is equivalent to 4(x2 + i) = (x + 2)2, that is
3x 2 - 4x + 4y2 = 4.
By completing the square this is the equation of an ellipse
29. ) x 2 + i = 4 - 2x. Since r G 0 by assumption, we must have 4 - 2x G 0, or equivalently x ~ 2. Conversely, for x ~ 2 the relation is equivalent with what we obtain when we square both sides, and the equation becomes
Al8 ANSWERS TO EXERCISES
or equivalently
3x2 - 16x - / = -16.
This is the equation of a hyperbola. Thus the equation in polar coordinates is equivalent with the equation of a hyperbola, together with the additional condition x ~ 2.
30. r = tan 8 = sin 8/cos 8. Multiply both sides by cos 8 to see that this equation is equivalent to r cos 8 = sin 8, that is
x = sin 8.
Of course, the function is not defined when cos 8 = O. Since
-1 < sin 8 < I,
it follows that -I < x < 1. We make a small table:
8 x
0 0 11/4 I/Ji 11/3 fi/2
inc. 0 to 11/2 inc. 0 to 1
As 8 approaches 11/2, x approaches 1. But cos 8 approaches 0 and so r = tan 8 becomes very large positive. Hence the graph looks as in the following figure for 0 ~ x < 1.
The graph also has a symmetry. Since r = y/x and r ~ 0, both y and x must have the same sign, that is both x, y > 0 or both x, y < 0, unless x = O.
ANSWERS TO EXERCISES
The next interval of (J for which tan (J is positive is then
3n n5:.(J<- . - 2
Either by making a table again, or by symmetry, using
tan(8 + n) = tan 8
you see that the graph is as shown for - I < x ~ o. 31. r = 5 + 2 sin 8. We make a table :
(J sin 8 r
o to n/2 inc. 0 to I inc. 5 to 7 n/2 to n dec. 1 to 0 dec. 7 to 5
n to 3n/2 dec. 0 to -1 dec. 5 to 3 3n/2 to 2n inc. -1 to 0 inc. 3 to 5
z=7
x =-5
A19
32. r = 11 + 2 cos (JI. Again we make a table. The absolute value sign makes the right-hand side positive, so we get a value of r for every value of 8. However, we want to choose intervals which take into account changes of sign of 1 + 2 cos 8, this is when cos 8 = -1/2. We make the table accordingly, when 8 = 2n/3 or 8 = 4n/3.
8 cos (J r
o to n/2 dec. 1 to 0 dec. 3 to 1 n/2 to 2n/3 dec. 0 to - 1/2 dec. 1 to 0
2n/3 to n dec. -1/2 to -I inc. 0 to 1
A20 ANSWERS TO EXERCISES
Since cos( - B) = cos B, the graph is symmetric with respect to the x-axis, and looks like this :
\
\ 2
-3 -2-1 3
33. (a) r = 2 + sin 28. Since sin 28 lies between - 1 and + 1. it follows that the right-hand side is positive for all B and so there is an r corresponding to every value of B. We make a table, choosing the intervals to reflect the regions of increase of sin 2B, so by intervals of length n/4.
B 2B sin 2B r
o to n/4 o to n/2 inc. 0 to 1 inc. 2 to 3 n/4 to n/2 n /2 to n dec. I to 0 dec. 3 to 2 n/2 to 3n/4 n to 3n/2 dec. 0 to -1 dec. 2 to 1
3n/4 to n 3n/2 to 2n inc. -I to 0 inc. 1 to 2
The graph looks like this.
33. (b) Make a table. We use intervals of length n/4 because sin B changes its behavior on intervals of length n/2, so sin 2B changes its behavior on intervals of length n/4.
ANSWERS TO EXERCISES A21
8 28 sin 20 r = 2 - sin 28
inc. 0 to n/4 inc. 0 to n/2 inc. 0 to 1 dec. 2 to 1 inc. n/4 to n/2 inc. n/2 to n dec. 1 to 0 inc. 1 to 2 inc. n/2 to 3n/4 inc. n to 3n/2 dec. 0 to -1 inc. 2 to 3 inc. 31t/4 to 1t inc. 31t/2 to 21t inc. -1 to 0 dec. 3 to 2
The part for 1t ;;; 8 ;;; 21t is obtained similarly, or by symmetry.
34. x ;;; 0, y = 0 (negative x-axis)
35. x = 0, y ~ 0 (positive y-axis)
36. x = 0, y ;;; 0 (negative y-axis)
37. 38.
39.
A22 ANSWERS TO EXERCISES
V, §I, p. 165
1. 1 2. ~ 3. ~ 4. 1 5. ~ 6. 0 7. ± 1
n 5n 8. 4 + 2nn and 4 + 2nn, n = integer. 9. nn, n = interger
n 10. "2 + nn, n = integer
V, §2, p. 175
1. Increasing for all x. 2. Decreasing for x ~ 1, increasing for x ~ 1. 3. Increasing all x.
4. Decreasing x ~ - j2i3 and x ~ j2i3. Increasing - j2i3 ~ x ~ j2i3. 5. Increasing all x. 6. Decreasing for x ~ O. Increasing for x ~ O.
17. Min: 1; max: 4 19. Max: -1; min: 4 21. Min: -1; max: -2, 24. Let f(x) = tan x-x. Then ['(x) = 1 + tan 2 x - 1 = tan 2 x. But tan 2 x is a
square, and so is > 0 for 0 < x < n/2. Hence f is strictly increasing. Since f(O) = 0 it follows that f(x) > 0 for all x with 0 < x < n/2.
26. Base = jC;3, height = jC/12 27. Radius = jC/3n, height = jC/3n
28. Base = JCi6, height = JCi6; Radius = jC/6n, height = 2jC/6n 30. (a) f(t) = -3t + C (b) f(t) = 2t + C 31. f(t) = -3t + 1 32. f(t) = 2t - 5 33. x(t) = 7t - 61 34. H(t) = - 2t + 30
VI, §1, p. 187
1 1 1. 0, 0 2. 0, 0 3. 0, 0 4. -, - 5. 0, 0 6. 00, - 00 7. - 00, 00
n n 8. -t, -t 9. -00, +00 10.0,0 11. 00, -00 12. -00, 00
13. 00, 00 14. - 00, - 00 15. 00, - 00 16. - 00, 00
17. 00, 00 18. - 00, - 00
19. n a. x -+ 00 x -+ -00
Odd >0 f(x) -+ 00 f(x) -+ - 00
Odd <0 f(x) -+ - 00 f(x) -+ 00
Even >0 f(x) -+ 00 f(x) -+ 00
Even <0 f(x) -+ - 00 f(x) -+ - 00
20. Suppose a polynomial has odd degree, say
f(x) = ax· + lower terms,
ANSWERS TO EXERCISES A23
and a =I- o. Suppose first a > o. If x ..... 00 then f(x) ..... 00 and in particular, f(x) > 0 for some x. If x ..... - 00 then f(x) ..... - 00, and in particular, f(x) < 0 for some x. By the intermediate value theorem, there is some number c such that fCc) = O. The same argument works if a < O.
VI, §2, p. 191
I. For sin x: 0, n, and add nn with any integer n.
n 3n 2. For cos x: 2' 2' and add nn.
3. Let f(x) = tan x. Then rex) = 1 + tan2 x, and
d rex) = dx (1 + tan2 x) = 2(tan x)(1 + tan2 x).
The expression 1 + tan2 x is always > 0, and rex) = 0 if and only if x = 0 (in the given interval). If x > 0 then tan x > 0 and if x < 0 then tan x < O. Hence x = 0 is the inflection point.
4. Sketch of graphs of sin2 x and Isin xl .
n n 3n n 3n 2n n n 3n 2n
4 2 4 2 2 2
Isin xl
Observe that the function sin2 x is differentiable, and its derivative is 0 at x = 0, n/2, n, etc., so the curve is flat at these points. On the other hand, Isin xl is not differentiable at 0, n, 2n, etc., where it is "pointed." For instance, the graph of Isin xl for n ~ x ~ 2n is obtained by reflecting the graph of sin x across the x-axis.
Let f(x) = sin2 x . Then rex) = 2 sin x cos x = sin 2x. Also rex) = 2 cos 2x. These allow you to find easily the regions of increase, decrease, and the inflection points, when rex) = O. The inflection points are when cos 2x = 0, that is 2x = n/2 + nn with an integer n, so x = n/4 + nn/2 with an integer n.
6. Bending up for x > 0; down for x < O.
7. Bending up for x > j3, - j3 < x ~ O. Down for x < - j3, o ~ x < j3.
8. Bending up for x> 1, -1 < x ~ O. Down for x < -1 and O~x<1.
9. Max at x = n/4. Min at x = Sn/4. Strictly increasing for 0 ~ x ~ n/4, Sn/4 ~ x ~ 8n/4. Decreasing for
n/4 ~ x ~ Sn/4.
Inflection points: 3n/4 and 7nj4.
A24 ANSWERS TO EXERCISES
Sketch of curve f(x) = sin x + cos x .
v'2
-1
-\1'2 (7 , -v'2)
VI, §3, p. 196
1. Let f(x) = ax3 + bx2 + ex + d. Then
f"(x) = 6ax + 2b.
There is a unique solution x = -b/3a. Furthermore, if a > 0:
f"(x) = 6ax + 2b > 0 ¢> x> -b/3a,
f"(x) = 6ax + 2b < 0 ¢> x < -b/3a.
Hence x = -b/3a is an inflection point, and is the only one. If a < 0, dividing an inequality by a changes the direction of the inequality, but the argument is the same.
17. (a), (e) 18. (c) 19. f"(x) = 12x2 + 18x - 2. So !f"(x ) = 6x 2 + 9x - I, and f"(x) = 0 if and only
if
-9-JiOs x = 12 or
-9+JiOs x = 12
Furthermore the graph of f" is a parabola bending up (because the coefficient 12 of x2 is positive) and so
f"(x) < 0 if and only if -9-JiOs -9+JiOs
12 < x < x = 12 .
graph of f"(x)
ANSWERS TO EXERCISES A25
Therefore rex) changes sign at the two roots of rex), whence these two roots are inflection points of f
We have f'(x) = 4x3 + 9x2 - 2x = x(4x2 + 9x - 2). The critical points of f are the roots of f', that is
-9-Jill X2 = 8 '
-9 + Jill X3 = 8 .
Note that f( - 2) < 0 (by direct calculation, so f(x) is negative at some x < O. Also
if x -+ - 00 then f(x) -+ 00,
if x -+ 00 then ((x) -+ 00 .
If x ~ 0 then f(x) > O. Indeed, if 0 < x ~ 1 then _x2 + 5 > 0 and the other two terms X4 + 3x3 are both positive, so f(x) > O. If x ~ I, then
and 3x 3 + 5 > 0 so again ((x) > O. We can now sketch the graph of f.
y-axis
graph off
inflection point
inflection point
-----..ll-+++-;-=:.-------x-axis
21. We have f'(x) = 6x s - 6x 3 + ~x = 3x(2x4 - 2X2 + t). Let u = x2. Then the roots of f'(x) (which are the critical points of f) are x = 0, and those values coming from the quadratic formula applied to u, namely
2u2 - 2u + ~ = O.
This yields u = 1/4 or u = 3/4, or in terms of x,
x = ± 1/2 and x = ±fi/2.
Observe that f(x) can also be written in terms of u = x 2, namely
A26 ANSWERS TO EXERCISES
Then you will find that the values of f at the critical points are all equal to 1/32 or -1/32. (Neat?) The graph looks like this.
y-axis
---+--1K-I---+-l-1H~-\-+---.I-t--- x-axis
The inflection points can also be found easily. We have
f"(x) = 30x4 - 18x2 + ~ = 30u2 - 18u + ~.
Hence f"(x) = 0 if and only if
-18 ± ji89 u = 60 .
This gives two values for u, and then x = ± Ju are the inflection points.
VI, §4, p. 201
Also see graphs below.
c.p. Increasing Decreasing
1. 3 ± J1i x ~ 3 - J1i and 3 - J1i ~ x < 3 and x~3+J1i 3<x~3+J1i
2_ 3±JW 3-JW~x~3+JW 3 + JW ~ x and x ~ 3 - flo
3. -I ± J2 -1-J2~x~ -I +J2 x ~ -I - J2 and x~-I+J2
For 4 and 5, see graphs below.
6·1 0 IX~O I J2 < x and 0 ~ x < /2 7_ None x< -1, X>-1
ANSWERS TO EXERCISES A27
ReI. ReI. Max. Min. Increasing Decreasing
9. -J3 J3 x < -J3, x> J3 -J3 <x <0, O<x<J3
12. None None Nowhere x < 5/3, x > 5/3 14. None 0 x>O -1<x<O
16. None None Nowhere x < -J5, -J5 < x < J5, x> J5
18. 0 None x < -2, 0<x<2, x>2 -2<x<0
4. y = f(x) = x - l/x; no critical point. Function strictly increasing on every interval where defined.
5. Y = f(x) = xl(x3 - I); f'(x) = -(2x3 + 1)/(x3 - 1)2
f"(x) = 6x 2(X 3 - 1)(x3 + 2)/(x3 - 1)4
= «x3 + 2)/(x 3 - I»· square.
A28 ANSWERS TO EXERCISES
Critical point at x = - 1/.y2. Inflection point at x = -.y2. Local max at critical point.
VI, §5, p. 211
1. rj2 by !rj2 where r is the radius of the semicircle. 2. Let x be the side of the base and y the other side, as shown on the figure.
y
Let A(x) be the combined area. Then
A(x) = x 2 + 4xy = 4S, whence
If the volume is V, then
Then V'(x) = 12 - 3x2/4 and V'(x) = 0 if and only if x = 4, and y = 2 by substituting back in the expression for y in terms of x. We have for x > 0:
V'(x) > 0 -= 12 - 3x2/4 > 0 -= x < 4,
V'(x) < 0 -= 12 - 3x2/4 < 0 -= x> 4.
Hence x = 4 is a maximum point. 4. The total cost f(x) is given by
f(x)=- 2+- -+D -= +-. 300 ( X2) 30 300 300D + ISO 3x x 600 100 x x 20
Then f'(x) = 0 if and only if x 2 = 20(300D + IS0)/3. Answers: (a) x = 20)3
(b) x = 40j2 (c), (d), (e) x = 60 because the critical points are larger than
60, namely 20ji3, 60j2 and 20)23 in the respective cases.
5. 4j2 by sj2 6. Answer: x = S/3. For x, y ~ 0 we require x + y = 4 and we want to mini
mize x 2 + y3. Since y = 4 - x we must minimize
Then f'(x) = 2x - 3(4 - xf, and f'(x) = 0 if and only if x = (26 ± 10)/6. The solution 36/6 is beyond the range 0 ~ x ~ 4. The critical point in this
ANSWERS TO EXERCISES A29
range is therefore x = 16/6 = 8/3. By a direction computation you can see that f(8/ 3) is smaller than f(O) or f(4). Since there is only one critical point in the given interval, it must be the required minimum. You could also graph f'(x) (parabola) to see that f'(x) < 0 if 0 < x < 8/3 and f'(x) > 0 if 8/3 < x < 4, so f(x) is decreasing to the left of the critical point, and increasing to the right of the critical point, whence the critical point is a minimum in the given interval.
7. Minimum: use 24n/(4 + n) cm for circle; 96/(4 + n) cm for square. Maximum: use whole wire for circle. We show how to set up Exercise 7. Let x be the side of the square. Then 4x is the perimeter of the square, and 0 ~ 4x ~ 24, so 0 ~ x ~ 6. Also, 24 - 4x is the length of the circle, i.e. the circumference. But
24 - 4n = 2nr, where r is the radius,
and nr2 is the area of the circle. The sum of the areas is
This is expressed in terms of two variables, x and r, but we have the relation
24 -4x r=~,
so that the sum of the areas can be expressed in terms of one variable,
( 24 - 4X)2 f(x) = n ~ + x 2.
You can now minimize or maximize by finding first the critical points, and then investigate if they are maxima, minima, or whether such extrema occur when 4x = 24 or 4x = O. Note that the graph of f is a parabola, bending up. What is f(O)? What is f(6)? The possible values for x are in the interval 0 ~ x ~ 6. The maximum or minimum of f in this interval is either f(O) , f(6), or at the critical point x such that f'(x) = O. Draw the graph of f to get the idea of what happens.
8. Answer: (1,2). The square of the distance of (x , y) to (2, 1) is
(x - 2)2 + (v - 1)2.
Since y2 = 4x, we get x = y2/4 and so we have to minimize
(l )2 f(y) = 4 - 2 + (y - 1)2.
But j'(y) = V /4) - 2 and f'(y) = 0 if and only if y = 2, so x = 1. You can verify for yourself that f'(y) > 0 if y > 2 and f'(y) < 0 if y < 2, so y = 2 is a minimum for f(y) .
9. (,fi /2, 1), (- ,fi/2, 1) 10. The square of the distance between (11, 1) and (x, y) is
(x - 11)2 + (y - 1)2 = (x - 11)2 + (x 3 - 3x - 1)2 = f(x),
A30 ANSWERS TO EXERCISES
which is a function of x alone. The picture is as shown roughly.
We have to minimize f(x). As x becomes large positive or negative, f(x) becomes large positive because
f(x) = x 6 + lower degree terms.
Hence a minimum occurs at a critical point. We have
f'(x) = 2(x - ll) + 2(x3 - 3x - 1)(3x2 - 3)
so
! f'(x) = 3x 5 - I2x 3 - 3x 2 + lOx - 8.
Plugging 2 directly shows that 1'(2) = 0, and therefore 2 is a critical point. If we could show that 2 is the only critical point of f, then we would be done. Let us get more information on f'(x) . By long division, we obtain a factoring
!f'(x) = (x - 2)g(x) where g(x) = 3x4 + 6x 3 - 3x + 4.
The function g(x) is still complicated. If g(x) * 0 for all x we would be done, but it looks hard, even if true. Let us avoid technical complications and let us simplify our original problem by inspection. The picture suggests that all points (x, y) on the curve y = x 3 - 3x such that x ~ I will be at further distance from (11, I) than (2,2). This is actually easily proved, for f(2) = 82, and if x ~ 1 then
f(x) ~ (x - 11)2 ~ 100 > 82.
Therefore it suffices to prove that f(2) is a minimum for f(x) when x > 1, and it suffices to prove that 2 is the only critical point of f(x) for x > I. Thus it suffices to prove that g(x) * 0 for x > I. This is easy because 3x4 + 4 is positive, and
6x 3 - 3x = 3(2x 3 - 1) > 0 for x> I.
ANSWERS TO EXERCISES A31
This proves that g(x) > 0 for x > 1. Therefore x = 2 is the only critical point of f for x > I, and finally we have proved that the minimum of f(x) is at x = 2.
11. (5, 3), (- 5, 3) 12. (-1/2, 1}2) 13. (-I, 0) 14. Answer (1,2). The square of the distance between (x, y) and (9,0) is
(x - 9)2 + y2. Since y = 2x2, we have to minimize
Then f'(x) = 16x3 + 2x - 18, and f'(l) = O. You can graph f'(x) as usual. Since f"(x) = 48x 2 + 2 > 0 for all x, we see that f'(x) is strictly increasing, so f'(x) = 0 only for x = I, which is the only critical point of f But f(x) becomes large when x becomes large positive or negative, and so f(x) has a minimum. Since there is only one critical point for f, the minimum is equal to the critical point, thus giving the answer.
15. F = 2}3 Q/9b2
16. Answer: y = 2h/3. We have F'(y) = y!<h - y)-1 /2( -1) + (h - y)I /2, so
, 2h - 3y F (y) = 2(h _ y)I /2 .
Thus F'(y) = 0 if and only if y = 2h/ l But F(O) = F(h) = 0 and F(y) > 0 for all y in the interval 0 < y < h. Hence the critical point must be the maximum.
17. (2, 0) L
18. max x = 0 (no triangle) ; min x = (I I)' Cut the wire with x de-2n - -+-
6}3 2n voted to the triangle and L - x to the circle. Let s be the side of the triangle so 3s = x; let h be the height of the triangle, and r the radius of the circle. Then
L = x + 2nr and
We need other relations to make A a function of one variable x, namely
h = s}3/ 2 = x}3/6 and r = (L - x)/2n.
Then A(x) = x 2 }3/36 + «L - x)/2n)2 and A'(x) = x}3/18 - (L - x)f2n. Thus the critical point is as given in the answer. Since the graph of A(x) is a parabola bending up, the critical point is a minimum. The maximum occurs at the end point of the interval 0 ~ x ~ L. To find out which end point, evaluate A(O) and A(L) and compare the two values to see that A(O) is bigger.
[ ( 13.5)2/3J3/2 13ji3 19. 4 I + 4 = -2-
A32 ANSWERS TO EXERCISES
20. We just set it up. Let r be the radius of the base of the cylinder, and h the height. Then the total volume, which is constant, is
This allows to solve for h, namely
h
The cost of material is a constant times:
(Area of cylinder) + (twice area of sphere).
We can express this cost as a function of rand h, namely
Since h is expressed as a function of r above, we get the total cost expressed as a function of r only, namely
[ (V - 4nr3/3) ] [2V 16 ] f(r) = C 2nr nr2 + 8nr2 = C ---; + 3 nr2 .
From here on, you can find f'(r) and proceed in the usual way to find when f'(r) = O. There is only one critical point, and f(r) becomes large when r approaches 0 or r becomes large. Hence the critical point is a minimum.
21. The picture is as follows :
The longest rod is that which will fit the minimal distance labeled z in the figure. We let x, y be as shown in the figure. We have
Z= J42+y2+~.
This depends on the two variables x, y. But we can find a relation between them by using similar triangles, namely
8 y
x 4'
so that
Hence
ANSWERS TO EXERCISES
32 Y=-·
x
z = f(x) = J16 + C:y + Jx2 + 64.
You then have to minimize f(x) . The answer comes out
x = 4,y4,
22. From the figure in the text, we have 0 ~ x ~ a, and
dist(P, R) = Jx2+YI, dist(R, Q) = J(a - X)2 + y~.
Hence the sum of the distances is
f(x) = Jx 2 + yi + J(a - X)2 + y~.
Then x (a - x)
f'(x) = - . Jx2 +yi J(a-x)2+y~
We have f'(x) = 0 if and only if
x a-x
Jx 2 +yi J(a-x)2+ yf which is the desired cosine relation.
A33
In particular, we see that there is only one critical point. Hence the minimum of f is either at the end points x = 0 or x = a, or at the critical point. We shall now prove that the critical point is the minimum. We are trying to prove that the graph of f looks something like this.
o critical point a
When x is near 0 and x > 0 then the first term in f'(x) is small and the second term is near
a
A34 ANSWERS TO EXERCISES
which is negative. Hence f'(x) < 0 when x is near 0, and the function is decreasing when x is near O.
On the other hand, suppose x is near a. Then the second term in f'(x) is near 0 and the first term is near
Consequently f'(x) > 0 when x is near a. Therefore f is increasing when x is near a. Since f(x) is decreasing for x near 0 and increasing for x near a it follows that the minimum cannot be at the end points, and hence is somewhere in the middle. Hence the minimum is a critical point, and we have seen that there is only one critical point. Hence the minimum is at the critical point. This proves what we wanted.
23. Let x and a - x be as on the figure. Then 0 ~ x ~ a.
YI
1 L----x ~----, l Y2
1 P2
Let tl be the time needed to travel from PI to Q, and t2 the time needed to travel from Q to P2 . Then
and
Then
Both VI' V2 are given as constant. Hence again we have to take f'(x) and set it equal to O. This is similar to Exercise 22, and we find exactly the relation that is to be proved.
24. p = sin. Since L(O) = L(I) = 0 and L(p) > 0 for 0 < p < I, it follows that the maximum is at a critical point. But
L'(p) = pS(n - s)(1 - pr s - I ( -1) + spS-I(1 - pr s
= pS - I(1 _ p)n-s-I[p(s - n) + s(l - p)]
= pS - I(1 _ p)n - S- I( _ np + s).
For 0 < p < I, the factor pS-I(1 - p)n-S-I is not 0, so L'(p) = 0 if and only if - np + s = 0, that is p = sin. Hence there is only one critical point, so the maximum is at the critical point.
25. (a)
ANSWERS TO EXERCISES A35
y-axis
--+---+---+--+--"'"----x-axis
Let x, y be the intercepts of the line with the axes. Then the area of the triangle is equal to
A = txy.
We want to minimize the area. By similar triangles, we know that
y
x x-3 so
x y= -- .
x-3
Then the area is given by
x 1 x 2
A(x)=tx - -= - - - . x-3 2x-3
From the physical considerations, we are limited to the interval x > 3. As x approaches 3 and x > 3 the denominator approaches 0 and is positive. Since x 2 approaches 9, it follows that A(x) becomes large positive. Also as x -> 00 , A(x) -> 00 . Hence the minimum of A will occur at a critical point.
We find the critical points. We have
, (x - 3)2x - x 2 x 2 - 6x 2A (x) = (x _ W = (x _ 3)2 '
Then A'(x) = 0 if and only if x = 6 (x = 0 is excluded because x> 3). Hence the desired line passes through the point (6,0). The equation is then
or also
1-0 y - 0 = -- (x - 6) = -1(X - 6).
3-6
y - 2 = -1X.
25. (b) 3y = -2x + 12 26. x = (a l + ... + an)/n. We are given
A36
Then
ANSWERS TO EXERCISES
f'(x) = 2(x - a\) + .... + 2(x - an)
= 2x - 2a\ + 2x - 2a2 + ... + 2x - 2an
= n2x - 2(a\ + . .. + an).
So f'(x) = 0 if and only if nx = a\ + .. . + an ' Divide by n to get the critical point of f. Since f(x) -+ 00 as x -+ ± 00 because for instance just one square term (x - a\)2 becomes large when x becomes large positive or negative, it follows that the minimum must be at a critical point, and there is only one critical point. Hence the critical point is the minimum.
27. 25/ fi 28. Answer: 8 = n/2. Let h be the height of the triangle as shown on the picture. _x_
It suffices to maximize the area given by
for 0 < x < 60.
Then A'(x) = (_2X2 + 602)/2(602 - X2)1 /2 and A'(x) = 0 if and only if
x = 60/ fi . But A(x) ~ 0 and A(O) = A(60) = O. Hence the critical point is the maximum. The above value for x implies that 8 = n/2, because the tri
angle is similar to the triangle with sides 1, 1, fi . 29. n/ l We work it out. Let the depth be y. Then sin 8 = y/ l00. Maximum
capacity occurs when the area of the cross section is maximum. This area is equal to
A = l00y + 2(h 100 cos 8).
You have a choice whether to express A entirely in terms of y or entirely in terms of 8. Suppose we do it in terms of 8. Then
A(8) = 104 sin 8 + 104 sin 8 cos 8 = 104 [sin 8 + ! sin 28].
So A'(8) = 104 [cos 8 + cos 28], and 0 ~ 8 ~ n/2. But in this interval, cos 8 and cos 28 are strictly decreasing so A'(8) is strictly decreasing. We have A'(8) = 0 precisely when cos 8 + cos 28 = 0, which occurs when 8 = n/l Thus A'(8) > 0 if 0 < 8 < n/3 and A'(8) < 0 if n/3 < 8 < n/2. Hence A(8) is increasing for 0 ~ 8 ~ n/3 and decreasing for n/3 ~ 8 ~ n/2. Hence the maximum o..:curs when 8 = n/3.
30. (a) a = 16 (b) a = -54. To see this, note that
f'(x) = 2x - a/x 2,
ANSWERS TO EXERCISES A37
and f'(x) = 0 if and only if a = 2x3• For x = 2 and x = -3, this gives the desired value for a. You can check that this is a minimum directly by determining when f'(x) > 0 or f'(x) < O. As for part (c), this is one of the rare cases when taking the second derivative is useful. The second derivative is f"(x) = 2 + 2a/x3, and the critical point has been determined to be when a = 2x3 ; so if x is the critical point we get f"(x) = 6 > O. Hence the critical point must be a local minimum.
c 31. 3 r:Jl. away from b
1 + ~a/b 32. base = 24/(4 + n) and height = 12/(4 + n')'). The picture is as follows.
x
nx perimeter = 12 = x + 2y + 2
Area = xy + !nr2 = xy + !n(x/2)2. You can solve for y in terms of x by using the perimeter equation, so the area is given as a function of x, A(x). Finding A'(x) = 0 yields the desired value for x. It is a critical point, the only critical point, and A(x) is a parabola which bends down, so the critical point is a maximum.
33. (a) Find the radius and angle of a circular sector of maximum area if the perimeter is 20 cm.
r
(J
r
Let r be the radius of the sector, and L the length of the circular arc. Then L = «(J/2n)2nr = (Jr. The perimeter is
P = 2r + L = 2r + (Jr = 20.
Hence we can solve for (J in terms of r, that is
20 (J = - - 2.
r
The area of the sector is
Graph of A(r)
r-axis
A38 ANSWERS TO EXERCISES
so in terms of r alone:
The graph of A(r) is a parabola bending down, so the maximum is at the critical point. But
A'(r) = 10 - 2r
so the maximum is when 2r = 10 so r = 5. Then
20 0=--2=2 5 .
33. (b) radius = 4 cm, angle = 2 radians 34. Don't make things more complicated than they need to be. Observe that
2 sin 0 cos 0 = sin 20. This is a maximum when 0 = n/4.
35. 2(1 + ~)312 36. r = (V/n)113 = y. Note that V = nr2y so y = V/nr2.
37.
y
---------
Let S = surface area so S = nr2 + 2nry. Then
S(r) = nr2 + 2V/r.
We have S'(r) = 2nr - 2V/r2 = 0 if and only if r = (V/n)113. There is only one critical point. But S(r) becomes large when r approaches 0 or also when r becomes large. There is a minimum since S(r) > 0 for r> 0, and so the minimum is equal to the single critical point.
P = 2r + L = 2r + rO.
From A = Or2/2 we get 0 = 2A/r2 so P can be expressed in terms of r only by
per) = 2r + 2A/r.
We have P'(r) = 2 - 2A/r2, and P'(r) = 0 if and only if r = A 1/2. So P has only one critical point, and per) -+ Cf) as r -+ Cf) and as r -+ O. Hence P has a minimum and that minimum, is at the critical point. This is a minimum for all values of r> O. In part (a), we have 0 ~ n so r2 ~ 2A/n, and the data limits us to the interval
ANSWERS TO EXERCISES A39
Hence in part (a) the minimum is at the critical point. In part (b), we have (J ~ n/2 so r2 ~ 4A/n, which limits us to the interval
Since J 4A/n > fl, the minimum is at the end point r = J 4A/n.
P(r) = 2r + 2A/r
38. x = 20. The profits are given by
P(x) = 50x - lex).
Then P'(x) = - 3x2 + 90x - 600. By the quadratic formula, P'(x) = 0 if and only if x = 20 or x = 10. But the graph of P(x) is that of a cubic, which you should know how to do. Also P(IO) is negative, and P(O) is also negative. So the maximum is at x = 20.
39. 18 units, daily profit $266
40. (50 + 5)94}/3. Same method as Problems 38 and 39. 41. 30 units; $8900 42. 20. Let q(x) be the profits. Since p = 1000 - lOx, we get
!lex) = x(lOOO - lOx) - f(x) = -20x2 + 800x - 6000.
The graph of g(x) is a parabola bending down, and g'(x) = 0 if and only if x = 20, which gives the maximum for q(x).
VII, §1, p. 221
1. Yes; all real numbers 3. Yes; all real numbers 5. Yes; for y < 1 7. Yes ; for y ~ 1 9. Yes ; for y ~ -1 11. Yes; for y ~ 2
13. Yes; for -1 ~ y ~ 1
VII, §2, p. 224
O. Let f(x) = _x 3 + 2x + 1. Then
f'(x) = -3x2 + 2
= 0 if and only if 3x2 = 2
if and only if x = J2j3 and x = - J2j3.
A40 ANSWERS TO EXERCISES
These are the critical points of.r. Also
f'(x) > 0 = -3x2 + 2 > 0
= 3x2 < 2
= Ixl < J2i3, f'(x) < 0 = x2 > 2/3,
= x> J2i3 and x < -J2i3. There are three maximal intervals where an inverse function of .r could be defined (excluding the end points):
x < -}2/3, x> j2;3.
Over each such interval, there is an inverse function of .r whose value at 2 is some point in the interval. We work out two out of the three cases, but you only had to pick one of them.
Case l. Observe that .r(l) = 2, and 1 > J2ii Therefore in this case, if 9 is the inverse function, then
1 1 g'(2) = f'(l)= _3+2=-1.
Case 2. Take the interval -J2i3 < x < J2/3. We want to solve .r(x) = 2, that is
- x3 + 2x + 1 = 2, or x 3 - 2x + 1 = O.
Factoring, this is the same as
(x - 1)(x2 + X - I) = o.
In the given interval, x = 1 is not a solution. There are two other possible solutions:
-I +J5 x=
2 and
-1- J5 x= 2
ANSWERS TO EXERCISES A41
But (- 1 - fl)/2 is not in the present interval of definition. Hence there is
only one possible x in this case, namely x I = ( - 1 + J5)/2. If 9 is the inverse function for the given interval, then
1 1 g'(2) = f'(x l ) = - 3x~ + 2·
(Alternative answers depend on the choice of intervals.)
I.! 2.h 3.!or-! 4.±1 S.±l 6.±!or±2~ 1 -I
7.! 8. -lor! ± fofl 9. -h 10. r::. or r::. 10",,2 10",,2
1 1 11. g'(y) = f'(x) = f'(g(y»'
-I -I g"(y) = f'(g(y»2 j"(g(y»g'(y) = f'(X)2 j"(x)g'(y).
If f'(x) > 0 then g'(y) > 0 by the first formula, and g"(y) < 0 by the second.
VII, §3, p. 229
and 2. View the cosine as defined on the interval
o ~ x ~ 1l.
On this interval, the cosine is strictly decreasing, and for 0 < x < 1l we have
d cos x ~= -sinx <0.
Hence the inverse function x = g(y) exists, and is called the arccosine. Since y = cos x decreases from 1 to - I, the arccosine is defined on the interval [-I, I]. Its derivative is given by g'(y) = I/f'(x), so that
d arccos y 1 - - - = g'(y) = -.-.
dy -SID X
But we have the relationship
sin 2 x = 1 - cos2 x,
and for 0 < x < 1l we have sin x > 0 so that
sin x = JI - cos2 x = JI-7.
A42 ANSWERS TO EXERCISES
-1 Consequently g'(y) = r.--::2.
Y 1- y2 The graph of arccos looks like this.
y-axis
1t Graph of y = arccos x
-----1---+-+----- x-axis -I
3. (a) 2/,fi (b) j2 (c) re/6 (d) re/4 (e) 2 (f) re/3
4. -2/,fi, -j2, re/3, re/4 5. Let y = sec x on interval 0 < x < re/2. Then x = arcsec y is defined on 1 < y,
I and dx/dy= ~.
yy y2 - 1
1 6. -re/2 7. 0 8. re/2 9. re/2 10. -re/4 11. 2x Jl - (x 2 - 1)2
-1 -1 4 12. 13. 14. ~ J _(x2 + 5x + 6) (arcsin X)2 jl=7 Y 1 - 4x2(arccos 2X)2
VII, §4, p. 233
1. 1T./4, re/6, -re/4, 1T./3 2. t, i, t,! 3. 1/(1 + i) 3 2cos2x
4. (a) -re/4 (b) 0 (c) -re/6 (d) re/6 5. 1 + 9x2 7. 0 9. 1 + sin2 2x
(cos x)(arcsin x) - (sin x)/ jl=7 -1 11. . 2 13. --2
(arcsin x) 1 + x
15. ~ (1 + arcsin 3X)2 17. (y - ~) = j2( x - ~)
19. (y - }) = ~ (x -f) 21. (y + i) = Js (x + D 22. 2/25 rad/sec
de 23. 440 ftjsec 24. 3/26 rad/sec 25. 0.02 rad/sec 27. de = ! sin e tan e
2 .J.. 28. M1 rad/sec 29. 82 rad/sec
25y 21 1500 1500
31. (a) (400)2 + (zp)2 rad/sec (b) (600)2 + (J.P)2 rad/sec
ANSWERS TO EXERCISES A43
33. Picture : rocket
x
We are given dx/dt = - 50. When t = 0 we have x(O) = 300, so in general
x(t) = 300 - SOt.
Furthermore, dy/dt = 3t2. We want to find dO/dt. We have
tan 0 = y/x so o = arctan y/x.
Then
dO x dy/dt - y dx/dt
dt - 1 + (Y/X)2 x 2
But y(5) = 125 and x(5) = 300 - 250 = 50. Hence
dO I 50·3·25 - 125 · ( - 50)
dt ,=5 = 1 + C52:Y 502
= 16/29 rad/see.
VIII, §1, p. 244
1. (a) y = 2e2x - e2 (b) y = 2e- 4x + 5e - 4 (c) y = 2x + 1 2. (a) y = !e- 2x + 3e - 2 (b) y = !e l /2x + !e l / 2 (c) y =!x + 1 3. y = 3e2x - 4e2 4. (a) esin 3x(eos 3x)3 (b) eos(eX + sin x)(eX + cos x)
(c) eos(ex+ 2)ex +2 (d) 4 eos(e4x-5)e4x - 5
1 . 5. (a) - - 2 eX (b) eX( -sm(3x + 5»)3 + eX eos(3x + 5)
1 + e X
. 1 (c) 2(eos 2x)eSln2x (d) - earccosx (e) _e - X
~
A44 ANSWERS TO EXERCISES
6. (c) Let f(x) = xex. Suppose you have already proved that
Then d pn+ l)(x) = - [(x + n)eX ]
dx
(derivative of a product)
= (x + n + 1 )ex•
This proves the formula for the (n + l)-th derivative. 8. (c) Differentiate f(x)leh(X) by the rule for quotients and the chain rule. We
get:
~ (f(X») = eh(X)f'(x) - f(x)eh(X)h'(x) dx eh(X) e2h(x)
eh(x)h'(x)f(x) - f(x)eh(X)h'(x) e2h(x)
=0.
Hence f(x)leh(X) is constant, so there is a constant C such that
f(x) eh(x) = C.
Now cross multiply to get f(x) = Ceh(x). 9. (y - e2 ) = 2e2(x - 1) 10. y - 2e2 = 3e2(x - 2) 11. y - 5e5 = 6e 5(x - 5)
12. y=x 13. (y-l)= -x 14. y-e- I =e-I(x-l) 15. Let f(x) = eX + x. Then f'(x) = eX + 1 > 0 for all x. Hence f is strictly in
creasing. We have f(O) = 1, and f( -1) = lie - 1 < 0 because lie < 1. By the intermediate value theorem, there exists some x such that f(x) = 0, and this value of x is unique because f is strictly increasing.
16. See the proof of Theorem 5.1. 17. If x = 1 in Exercise 16(b) we get 2 < e. If x = 1 in Exercise 16(c) we get
2.5 < e. 18. See Theorem 5.2. 19. (a) Let fl(x) = e - x - (1 - x). Then f'1(X) = _e- x + 1 and since eX> 1 for
x> 0 we get
for x> O.
Hence f is strictly increasing for x ~ O. Since fl(O) = 0 we conclude fl (x) > 0 for x > 0, in other words
e- X - (l - x) > 0 for x> 0,
and therefore e - X > I - x for x > 0, as desired.
ANSWERS TO EXERCISES A45
f~(x) = -I + x + e- X = fl(x) .
By part (a), we know that fl(x) > 0 for x > O. Hence f2 is strictly increasing for x ~ O. Since f2(0) = 0 we conclude that f2(x) > 0 for x> 0, whence (b) follows at once.
( X2 x3 )
(c) Let f3(x) = e- x - 1 - x + "2 - N . Then flex) = f2(x) . Use part (b)
and similar arguments as before to conclude f3(x) > 0 for x> 0, whence (c) follows.
(d) Left to you.
20. If we put x = 1/2 in Exercise 19(a), then we find t < e - I / 2 , or in other words, t < l /e l /2 . Hence el /2 < 2 and e < 4. If we put x = 1 in Exercise 19(c), then we find
that is
whence e < 3.
21. cosh2(t) = Het + e- t)2 = !(e2t + 2ete- t + e- 2t )
= He2t + 2 + e- 2t ).
22.
Similarly,
Subtracting yields cosh 2 - sinh2 = 1. As for the derivative, cosh'(t) = t(et - e- t); don't forget how to use the
chain rule : let u = -to Then
de- t de" du - =- - = _e- t
dt du dt .
y-axis
y = cosh x = e" + e- ' 2
y-axis
. e" - e- ' y = smh x = --2-
A46 ANSWERS TO EXERCISES
VIII, §2, p. 255
I. (a) (y - log 2) = !(x - 2) (b) (y - log 5) = *<x - 5) (c) (y - log!) = 2(x - !)
2. (a) (y - log 2) = -(x + 1) (b) (y - log 5) = ~(x - 2)
-3 (c) (y - log 10) = 5 (x + 3)
cos xli 3. (a) - . - (b) cos(log(2x + 3») -2 -·2 (c) --·2x
Sill x x + 3 x2 + 5
(sin x)/ x - (log 2x) cos x (d) . 2
Sill X
1 4. (y -log 4) = l(x - 3) 5. (y - log 3) = ~(x - 4) 7. (y - 1) = - (x - e)
e
8. (y - e) = 2(x - e) 9. (y - 2 log 2) = (1 + log 2)(x - 2)
3 -1 10. (y - 3) = - (x - e) 11. (y - 1) = - (x - e)
e e
12. (y __ 1_) = - I (x _ 2) log 2 2(1og 2)2
2x -I log x-I 14. -2-- 15 16 17. Wog X)-2/3 + (log X)1 /3
X + 3 . x(log X)2 • (log x)2
-x 18. - - 2
I - x
19. Let f(x) = x + log x. Then rex) = 1 + I/x > 0 for x > O. Hence f is strictly increasing. Also rex) = _1/X2, so f is bending down. Note that f(l) = 1. If x -. 00 then f(x) -. 00 because both x and log x become large. In fact f(x) lies at a distance log x above the line y = x. As x -. 0, log x -. - 00
(think of x = I/ez = e- Z where z -. 00). Hence the graph looks like this:
graph of f(x) = x + log x
x
ANSWERS TO EXERCISES A47
VIII, §3, p. 261
1. lOX log 10, 7X log 7 2. 3x log 3, rrX log rr 5. (y - 1) = (log 10)x 6. y - rr2 = rr2 log rr(x - 2) 7. (a) e" I08X[log x + I] (b) x(XX'[xx- 1 + (log x)xx(l + log x)] In (b), we write
and use the chain rule. The derivative of (XX) log x is found by the rule for differentiating a product, and we have
The derivative of XX was found in (a), and the answer drops out. S. (a) y - 1 = x-I (b) y - 4 = 2(1 + log 2)(x - 2)
(c) y - 27 = 27(1 + log 3)(x - 3) 9. (a) Let J(x) = XXI/2 = exl/2lolX. Then
f'(x) = x..rx[_I- + _1_ log xJ Jx 2Jx
so
1'(2) = 2..'2[_1 + _1_ log 2J. }2 2}2
The tangent line at x = 2 is
y _ 2..'2 = 2..'2 [ _1_ + _ 1_ log 2J(x - 2). }2 2}2
(b) y - 5v's = 5v's[_1_ + _1_ log 5J(X - 5) Js 2Js
10. Let J(x) = xlf< = ex l/ 3logx. Then
rex) = xvx[ x 1 /3 ~ + tx - 2/3 log x J (a) Tangent line at x = 2 is
y - 221 / 3 = 22I iJ [2-2 /3 + t2- 213 log 2](x - 2).
(b) Tangent line at x = 5 is
y - 551iJ = 551 /'[5 - 2/3 + !5- 2/ 3 10g 5](x - 5).
11. Let J(x) = x· - 1 - a(x - 1). Then J(I) = 0, and
rex) = ax·- I - a.
A48 ANSWERS TO EXERCISES
If x > I, then f'(x) > 0 so f(x) is strictly increasing. If x < 1 then f'(x) < 0 so f(x) is strictly decreasing. Hence f(l) is as minimum value, so that for all x> 0 and x'# 1 we get f(x) > O.
12. x = 0 and x = - 2/log a
13. We have (1 + ~r = (1 + ~r· If y -+ 00, then (1 +~)' approaches e, by
Limit 3, so its r-th power approaches e. This uses the fact that the r-th power function is continuous. If z approaches zo, then z' approaches zO. Here
and z approaches e by Limit 3. ah - 1 f(h) - f(O)
14. (a) We can write -h- = h where f(x) = aX. Hence the limit is
equal to 1'(0). Since I'(x) = aX log a we find 1'(0) = log a.
ah - 1 (b) Putting h = l /n we have n(a l /- - 1) = -h- so the limit comes from part
(a).
VIII, §4, p. 266
1. -log 25 2. 5e- 4 3. e-(IogI0)IO - 6, 4. 20/e 5. -(log 2)/K 6. (log 3)/4
7. 12 log IO/Iog 2 8. IOg~3~~:: 10 10. 1984: (50,000)284/50; 2000: 2 = 105
11. 30[~]S/ 3 12. 4[IOg -toJ 13. 2[IOg iJ S log! log!
14. (a) 40[IOg ioJ (b) 40[IOg 15J (c) 100[1]1 /2 15. log 2 log ~ log ~ S
16. (a) Sl68 log ! (b) 5568(log 4/5)/(Iog 1/2)
VIII, §5, p. 274
2. 3.
(-I, -l ifo ) fi2
ANSWERS TO EXERCISES A49
4. 5.
(-I, l i e) (I, I je)
-I
6. 7.
2
f(x) = e'/x
8.
3
f(x) = e"/x3
A50 ANSWERS TO EXERCISES
9.
,
11.
/ /
!(x) = e' - x
!(x) = e- X + x
10.
!(x) = eX + x
/ 12.
fix) = x - log x
13. Suppose first a < O. Let f(x) = eX - ax. When x is large positive, then
is large positive. When x is large negative, then eX is close to 0 and - ax is large negative, so f(x) is negative. By the intermediate value theorem, the equation f(x) = 0 has a solution.
Suppose next that a ~ e. Then f(l) = e - a ;£ 0, and again f(x) is large when x is large positive. The intermediate value theorem again provides a solution.
n 14. (a) - 2n log 2
(b) Limit is 0 in both cases. For instance, let x = e - Y. As x approaches 0, then y becomes large, and
xlogx= -ye- Y = -~, e
which approaches 0 by Theorem 5.1. Also x2 10g x = x(x log x), and the product of the limits is the limit of the product of x and x log x, so is equal to O.
ANSWERS TO EXERCISES
15. Let x = e- Y• Then log x = -y and
As x ~ 0, y ~ 00 so x(log x)" ~ 0 by Theorem 5.1. 16. Let x = eY• As x ~ 00, y ~ 00 so
(log x)" y" ---=-~O
x eY by Theorem 5.1.
17. (a) (b) !(x) = Xl log X
(l i e, -l ie)
(c) (d)
e
lie e
17. (a) f(x) = x log x, defined for x> O. Then:
We have:
1 f'(x) = x· - + log x = 1 + log x.
x
f'(x) = 0 ¢> log x = -I
f'(x) > 0 ¢> log x > - 1
f'(x) < 0 ¢> log x < -1
¢>
¢>
¢>
x = e- 1,
x>e-l,
x < e- 1•
AS1
So there is only one critical point, and the regions of increase and decrease are given by the regions where f' > 0 and f' < O.
We also get f"(x) = I/x > 0 for all x > 0, so f is bending up.
If x ~ 00 then log x ~ 00 also, so f(x) ~ 00.
If x ~ 0 then f(x) ~ 0 by EX,ercise 14(b).
This justifies all the items in the graph.
A52 ANSWERS TO EXERCISES
17. (b) We carry out the details of the graph for f(x) = Xl log x. We have
rex) = x + 2x log x = x(l + 2 log x).
Since x > 0, it follows that rex) > 0 if and only if 1 + 210g x > 0, and
1 + 2 log x> 0 if and only if
if and only if
logx> -1/2
Thus f is strictly increasing for x ~ e- 1/2 and is strictly decreasing for
From Exercise 14 we know that x log x approaches 0 as x approaches O. Hence rex) approaches 0 as x approaches 0, which means the curve looks flat near O. We have
f"(x) = (1 + 2 log x) + 2 = 3 + 2 log x.
Then f"(x) = 0 if and only if 3 + 210g x = 0, or, in other words, log x = - 3/2 and x = e - 3/2. Thus the inflection point occurs for x = e- 3/ 2• This explains all the indicated features of the graph.
17. (c) f(x) = x(log X)2 for x> O. Then
1 rex) = X· 2(log x) - + (log X)2
X
= (log x)(2 + log x).
The signs of the two factors log x and 2 + log x will vary according to the intervals when either factor is O. We have
rex) = 0 -= log x = 0 or 2 + log x = 0
-= x = 1 or log x = - 2
-= x=1 or x=e- 2 .
So there are two critical points at x = 1 and x = e- 2 . We make a table of regions of increase and decrease corresponding to the intervals between critical points.
interval log x 2 + log x rex) f
0<x<e- 2 neg. neg. pos. s.i. e- 2 <x<1 neg. pos. neg. s.d.
1 < x pos. pos. pos. s.i.
ANSWERS TO EXERCISES A53
The second derivative is not too bad:
1 1 2 f"(x) = (log x) - + - (2 + log x) = - (1 + log x).
x x x
For x> 0 we have 2/x > 0 so f"(x) = 0 if and only if x = e- I . Since
f"(x»O ¢> logx> -1 ¢> x>e - I ,
f"(x)<O ¢> logx< -1 ¢> x<e- I ,
it follows that x = e - 1 is an inflection point. The graph bends up if x > e - 1 and bends down if x < e - I.
If x -+ 00 then log x -+ 00 so f(x) -+ 00.
If x -+ 0 then f(x) -+ 0 by Exercise 15.
This justifies all features of the graph as drawn. 17. (d) We carry out the details. Let f(x) = xflogx for x> 0, and x#- 1. Then
( 1) log x-I f'(x) = log x - X· ~ /(log X)2 = (log X)2 .
For x#- 1 the denominator is positive (being a square). Hence
f'(x) = 0 ¢> log x = 1 ¢> x = e,
f'(x) > 0 ¢> log x > 1 ¢> x > e,
f'(x) < 0 ¢> log x < 1 ¢> x < e.
Next we list the behavior as x becomes large positive, x approaches 0, and x approaches 1 (since the denominator is not defined at x = 1).
If x -+ 00 then x/log x -+ 00 by Theorem 4.3.
If x -+ 0, then x/log x -+ O.
This is because log x becomes large negative, but is in the denominator, so dividing by a large negative number contributes to the fraction approaching O.
If x -+ 1 and x > 1 then x/log x -+ 00. Proof: The numerator x approaches 1. The denominator log x approaches 0, and is positive for x > 1. So x/log x -+ 00.
If x -+ I and x < I then xflog x -+ - 00 . Proof: Again the numerator x approaches 1, and the denominator log x approaches 0 but is negative for x < 1, so x/log x -+ - 00.
This already justifies the graph as drawn in so far as regions of increase and decrease are concerned, and for the critical point (there is only one critical point). Let us now look at the regions of bending up
A54 ANSWERS TO EXERCISES
and down. We write the first derivative in the form
Then
Therefore:
1 1 f'(x)= - - --.
log x (log X)2
-I 1 1 rex) = -1- - 2 - - (-2)(log X)-3-
(og x) x x
-I 1 = (log X)3 ~ (log X - 2).
rex) = 0 <0> log x = 2 <0> X = e2.
We shall now analyze the sign of rex) in various intervals, taken between the points 0, 1, and e2 which are the points where the factors of rex) change sign. Note that the sign of rex) (plus or minus) will be determined by the signs of log x, x, and log x - 2, together with the minus sign in front.
If x> e2 then rex) < 0 and the graph bends down, because log x - 2 > 0, both log x and x are positive, and the minus sign in front makes rex) negative.
If 1 < x < e2 then rex) > 0 because log x - 2 < 0, both log x and x are positive, and the minus sign in front together with the fact that log x - 2 is negative make rex) positive. Hence the graph bends up for I < x < e2.
If 0 < x < 1 then rex) < 0 because x is positive, log x is negative, log x - 2 is negative, and the minus sign in front combines with the other signs to make rex) negative. Hence the graph bends down for O<x<1.
18. Let f(x) = XX = exlogx. Then
f'(x) = eX log x( X· ~ + log x) = exlogx(l + log x).
Since e" > 0 for all numbers u, we have exlogx > 0 for all x> 0, so
f'(x)=O <0> logx=-1 <0> x=e- I,
f'(x»O <0> logx>-1 <0> x>e- I.
This already takes care of Exercise 18. 19. Note that f(e-I) = (e-I)I /e = e-I /e = I/el /e. Also
f'(x)<O <0> logx<-l <0> x<e- I.
Finally rex) = eX log x[1 /x + (1 + log X)2] > 0 for all x > 0 so the graph bends up.
ANSWERS TO EXERCISES ASS
f(x) = x' = e· I ....
l ie
20. Note that x-x = I/xx. But you should go through the rigamarole about taking the derivative etc. to see the graph as follows.
e
lie
f'(x) = ex(log 2+ loex{ X· ~ + log 2 + log x]
= 2XxX(1 + log 2 + log x).
Since 2'xx > 0 for all x > 0, we get
f'(x) > 0 = 1 + log 2 + log x > 0
= log x > - I - log 2
= x> e-I-log2
and e - 1 - 1og 2 = e-1e- 1og 2 = 1/2e, as was to be shown. 22. (a) I (b) I (c) lie (d) I
IX, §1, p. 291
sin 3x 1. -(cos 2x)/2 2. -3- 3. log(x + I), x> -I 4. log(x + 2), x> -2
IX, §3, p. 296
1. 156 2. 2 3. 2 4. log 2 5. log 3 6. ~ 7. e - I
A56 ANSWERS TO EXERCISES
IX, §4, p. 307
1. (a) U~ = ~G - 1) + 4(2 -~) L~ = 1(~ - 1) + ~(2 - ~)
(b) U~ = -'I(~ - 1) + 1f(i - ~) + 4(2 - i) Lf = 1(~ - 1) + -'I<i - ~) + 1f(2 - i)
(c) u~ = ti(i - 1) + ~G - i) + ~G - ~) + 4(2 - t) L~ = l(i - 1) + tiG - i) + ~G - ~) + ~(2 - t)
(d) U~=~-[(I+~r +(I+~r + ... +(I+~rJ 1 [ (1)2 (n - 1)2J L~ = ~ 1 + 1 + ~ + ... + 1 + -n-
2. (a) U~ = IG - 1) + ~(2 -~) + t(t - 2) + %(3 - t) L~ = ~(~ - 1) + t(2 - ~) + ~(t - 2) + i(3 - t)
(b) U~ = l(~ - 1) + ~(i - ~) + ~(~ - i) + iG - ~) + ~(~ -1) + ~(! -~)
L~ = i(~ - 1) + ~(i -~) + i(~ - i) + ~G -~) + ~(~ -1)+ ~(~ -~) (c) u~ = l(i - ~) + ~(i - i) + ~G - i) + 1(£ - t)
+ ~(~ - £) + Wi! - ~) + -to(~1 -1.p) + net - ¥) L~ = ~(i - ~) + ~(i - i) + 1G - i) + ~(~ - i)
+ ~(~ - £) + 140(1.p - ~) + n(¥ - 14°) + nei - ¥)
3. (a) U5 = t(f - 0) + 1(1 - f) + H~ - I) + 2(2 - ~) L5 = O(f - 0) + to - t) + IG - 1) + ~(2 - ~)
(b) U~ = Hi - 0) + H~ - i) + I(~ -~) + ~(~ - 1) + i(i - ~) + 2(~ - i)
L5 = O(i - 0) + t(~ - i) + ~<i - ~) + I(~ - 1) + ~(i - ~) + i(~ - i)
(c) U5 = Hi - 0) + t(f - i) + ~<i - f) + 1(1 - i) + i(i - I) + ~G - i) + iG -1) + 2(2 - i)
L5 = O(i - 0) + i(t - i) + f<i - f) + W - i) + l(i - 1) + iG - i) + 1(i -1) + ;t<2 - i)
(d) U5 =! [! + ~ + .. . + 2nJ L2 =! [0 +! + . .. + 2n - IJ nnn nOn n n
ANSWERS TO EXERCISES
4. (a) U5 = Ht - 0) + 1(1 - t) + £(~ - 1) + 4(2 - ~) L~ = oct - 0) + !(1 - t) + IG - 1) + £(2 - ~)
(b) U~ = ~(! - 0) + ~(~ -!) + 1(1 - ~) + -';(~ - 1) + ~(i - ~) + 4(2 - i)
L~ = O(! - 0) + ~ (~ - !) + W - ~) + 1(~ - 1)
+ -';(i - ~) + ~(2 - i) (c) U5 = he! - 0) + Ht - !) + h(~ - t) + 1(1 - ~)
+ H(~ - 1) + £G -~) + tiG -~) + 4(2 -~) L5 = O(! - 0) + he! - i) + i(~ - !) + h(l - ~)
+ l(~ - 1) + HG -~) + £G -~) + ti(2 -~)
(d) U~ = ~ [ :2 + (~r + ... + e:r] L~ = ~ [0 + n12 + ... + en: lr]
2 1[ 1 1 1 5. U I = ~ 1 + -( 1) + ... + ( n _ 1) 1 + ~ 1 + -n-
= [~+_1_ + .. . + _1_J n n + 1 2n - 1
L; ~~ k: ~) + (I : ~) + + (I: ~)l = [ _ 1_ + _ 1_+ ... + ~J
n+l n+2 2n
A57
6. The area under the curve y = l / x between x = 1 and x = 2 is log 2 - log 1 = log 2. Write down that this area is less than an upper sum and greater than a lower sum, and use Exercise 5 to get the desired inequalities.
7. U~ = [log 2 + log 3 + . .. + log n] = log n! L~ = [log I + log 2 + . . . + log(n - 1)] = log(n - 1)!
X, §1, p. 317
1. ¥ 2. 0 3. 0 4. 0
5. (b) and (c) Let f(x)=X- I / 2 , and let the partition be P= {1,2,3, ... ,n} for the interval [1, n]. Then compare the integral
fn xi/lin I f(x) dx = 172 I = 2(Jn - 1)
with the upper and lower sums.
6. (a) Use f(x) = x 2 and P = {O, 2, . .. ,n}, with interval [0, n]. The lower sum could start with 0, but this 0 may be omitted since 0 + A = A for all numbers A. (b) Let f(x) = x 3, interval [0, n], partition P = {O, . . ,n}. (c) Let f(x) = XI /4 , interval [0, n], partition P = {O, . .. ,n}.
AS8 ANSWERS TO EXERCISES
7. (d) Use J(x) = l/x4 over the interval [1, n] with the partition p, .. ,n} consisting of the positive integers from 1 to n. Then
Comparing with the lower and upper sum yields
11 11 I I 1 - + - + ... + - S; - - - S; 1 + - + .. . + ---. 24 34 n4 - 3 3n 3 - 24 (n - 1)4
8. Let J(x) = _1_2, Then fl J(x) dx = arctan x II = n/4. On the interval I + x 0 0
[0, I] use the partition P = {o), ~, ... ,~}. Draw the picture. n n n
9. Let J(x) = x 2• Then fol J(x) dx = ~31: = 1/3. On the interval [0, 1] use the
partition P = {O, l /n, ... ,n/n}. Draw the picture.
10. f log x dx = (x log x - x) I~ = n log n - n + I.
The lower sum is log I + log 2 + ... + log(n - 1) = log(n - I)! and
elowersum = (n - I)!, because e1og " = u.
On the other hand,
Since lower sum ~ integral, we get elower sum ~ einlegral, so
The upper sum is log 2 + log 3 + ... + log n = log n! and so eupper sum = n! Since integral ~ upper sum, we have einlegral ~ euppersum. Since the integral is n log n - n + 1, we get
X, §2, p.325
1. X4 2. 3x5/5 - x 6/6 3. -2 cos x + 3 sin x 4. ~X5/3 + 5 sin x 5. 5eX + log x 6. 0 7. 0 8. e2 - e - I 9. 4· 28/3
ANSWERS TO EXERCISES
10. In this problem, the curves intersect at x = 0 and x = 1.]
Hence the area is
11. ! from - 1 to 1 12.-A 13.! 14. i + ,52'
15. fi - 1. In this problem the graph is as follows.
-4 2
A59
The first point of intersection is when x = n14. Hence the area between the curves is
f"4
o (cos X - sin x) dx.
16. 9/2 17. t -! 18. 0 19. n2/2 - 2 20. 4 21. 4 22. 1 23. 4 24. _n2
25. 4 26. (a) 14 (b) 14 (c) 2n
X, §4, p. 334
1. Yes, j2 2. No 3. Yes, n12. We have
fB 1 IB -1 --2 dx = arctan x = arctan B - arctan 0 o + x 0
= arctan B.
A60 ANSWERS TO EXERCISES
As B -+ 00, arctan B -+ 1[/2. Remember the graph of arctan x which is as follows.
2
4. No. Let 0 < b < 5. Then
f: 5 ~ x dx = -log(5 - X)I: = -[log(5 - b) -log 5]
= log 5 -log(5 - b).
As b approaches 5, 5 - b approaches 0, and log(5 - b) becomes large negative. Hence the integral from 0 to 5 does not exist.
5. Let 2 < a < 3. Then
f3 1 13 Q X _ 2 dx = log(x - 2) a = log 1 - log(a - 2)
= -Iog(a - 2).
As a -+ 2, log(a - 2) -+ - 00, so -Iog(a - 2) -+ 00 and the integral from 2 to 3 does not exist.
6. No. 7. Here we have 0 ~ x < 2. Remember that we have the indefinite integral.
f ~ dx = log( - x) if x < o.
Therefore, if x < 2 then
f x ~ 2 dx = log(2 - x).
You can verify this by the chain rule, differentiating the right-hand side. Don't forget the -1. Note that when x < 2 then 2 - x > O. The log is not defined at negative numbers. Now you can find the definite integral for o < a < 2. We have
fa 1 la o x _ 2 dx = log(2 - x) 0 = log(2 - a) - log 2.
ANSWERS TO EXERCISES
The right-hand side approaches - 00 as a -+ 2, so the integral
-- dx= lim - - dx f 2 1 fO 1 o x - 2 0-2 0 X - 2
does not exist. 8. Improper integral does not exist. Let 2 < c < 3. Then
I3--I --o-2dx = I3(X - 2)-2dx = -(x _ 2)_113
.~-~ . . = _[1 __ 1 J.
c-2
As c -+ 2, the quotient 1/(c - 2) becomes arbitrarily large. 9. Does not exist.
10. Exists. Let 1 < c < 4. Then
f(X - 1)-213 dx = 3(x - 1)1 /31~ = 3[31 /3 - (c - 1)113].
As c -+ I, (c - 1)1/3 -+ 0, and so
lim 14 (x - 1)- 2/3 dx = 3.3 1/3. c-l c
11. Integral exists for s < 1. Let 0 < a < 1. Then for s '" 1, we get
x dx=--- =-----. II -s x- s + 1 II 1 al - s
a -s + 1 a 1 - s 1 - s
A61
If s < I, then 1 - s > 0 and a l - s approaches 0 as a approaches O. Hence the limit of the right-hand side as a -+ 0 exists and is equal to 1/(1 - s). On the other hand, if s > 1, then
1 a l - s = -
a'-I'
and s - 1 > 0, so as-I -+ 0 as a -+ 0, and I /as - I -+ 00, so the integral from 0 to I does not exist. When s = 1,
- dx = log x = log 1 - log a = -log a. II 1 II a X 0
When a -+ 0, log a -+ - 00 so the integral does not exist.
12. The integral Sf x- s dx exists for s> 1 and does not exist for s < 1. Evaluate
x- s dx = -- - --. fB 8-s+1 1
I l-s l-s
A62 ANSWERS TO EXERCISES
If 5> 1 then we write B- 5 + 1 = I /B5 - 1 and 5 - 1 > 0, so I /B5 - 1 ~O as B ~ 00. The limit of the integral exists and is equal to 1/ (5 - 1). If 5 < I, then B- 5 + 1 = B I - 5 becomes arbitrarily large when B becomes large.
13. Yes. Let B > 1. Then
As B becomes large, l/eB approaches 0, and the integral from 1 to B approaches l /e. It exists.
14. Does not exist. 15. -te- 2B + te- 4 . Yes, te- 4.
XI, §l, p. 339
I. eX ' /2 2. _ie-x' 3. Ml + X 3 )2 4. (log x)2/2
(Iogx)-n+1 5. 1 _ n if n =j:. I, and log(logx) if n = 1. 6. log(x2 + X + 1)
sin2 x sin3 x 2 7. x - log(x + 1) 8. - 2- 9. - 3- 10. 0 II. 5 12. -arctan(cos x)
13. Harctan X)2 14. 2/ 15. Let u = 1 - x 2. 15. -i cos(n2/2) + i 16. (a) -(cos 2x)/2 (b) (sin 2x)/2 (c) -(cos 3x)/3 (d) (sin 3x)/3 (e) e4x/4
(f) eSx/5 (g) _e- sx/5 17. -te- B' + t· Yes, t 18. _ie- BJ + i · Yes, i
XI, §l, Supplementary Exercises, p. 340
1 4 sins x ~ - 1 -1 I. 4log(x + 2) 3. -5- 5. yX -1 7. 6(3x2 + 5) 9. 2sin2 x
1 [U I7 / S U I2IS] -cos 3x II. 3 17/5 - 12/5 where u = x 3 + 1. 13. 3 15. -cos eX
4)2 2 n n2 1 17. log(log x) 19. log(eX + 1) 21. i 23. - 3- - 3 25. 4 27. 72 29. e - ~
XI, §2, p. 344
I. x arcsin x +~. Let
u = arcsin x,
du= ~dx, y 1- x 2
dv = dx,
v = x.
ANSWERS TO EXERCISES A63
Then
f arcsin x dx = f u dv = x arcsin x - f h dx.
Now the integral fh dx can be done by substitution, letting 1 - x 2
U = 1 - x2 and du = - 2x dx, so
f ---c:=x== dx = _- _1 f U- 1/2 duo ~ 2
2. x arctan x - ! log(x2 + 1). Let u = arctan X.
3. ~2; (2 sin 3x - 3 cos 3x). Let I = f e2x sin 3x dx. Let
u = e2x, dv = sin 3x dx,
du = 2e2x dx, v = -t cos 3x.
Then
1= -te2x cos 3x + ~ f e2x cos 3x dx.
Apply the same procedure to this second integral, with
u = e2x, dv = cos 3x dx,
du = 2e2x dx, v = t sin 3x.
Then
I = - te2x cos 3x + ~Ue2x sin 3x - ~I]
= -te2x cos 3x + ~e2x sin 3x - ~I.
Thus we see I appearing on the right-hand side with some constant factor. We can solve for I to get
¥I = ~e2X sin 3x - te2x cos 3x.
Multiply by 9 and divide by 13 to get the answer.
4. roe-4x sin 2x - !e- 4x cos 2x
A64 ANSWERS TO EXERCISES
5. x(log X)2 - 2x log x + 2x. Let u = (log X)2, so
dv = dx, v = x. Then
I du = 2(log x) - dx. Let
x
f (log X)2 dx = x(log X)2 - f ~ log x dx.
Then x/x = I, and you are reduced to f log x dx = x log x-x. 6. (log X)3 X - 3 f (log X)2 dx 7. X2~ - 2xex + 2ex
8. _x2e- x - 2xe- X - 2e- x 9. -x cos x + sin x 10. x sin x + cos x 11. - x2 cos X + 2 f x cos x dx 12. x 2 sin x - 2 f x sin x dx 13. Hx2 sin x2 + cos x 2]. Write
First let u = x2, du = 2xdx and use substitution to get
I = ~ f u cos u duo
Then use integration by parts. 14. -HI - X2)3/2 + ~(l - X 2)5/2 - ~(I - x2)7!2. Let u = I - x 2:
f x 5JI=7 dx = f x4JI=7xdx
= - ~ f (I - U)2U 1/2 du
= - ~ f (U 1/2 - 2U 3/2 + US/2) du
15.1X3Iogx-!x3. Let u=logx and dv=x 2dx. Then du=(I /x)dx and v=x 3/3 so
f x2 log x dx = l x3 log x - ~ f x2 dx.
X4 X4
16. (log x) '4-16
17. (log X)2 x33
- ~ f x2 10g x dx. Repeat the procedure to get the complete
answer.
18. - tx2e- x2 - te- x2 • Let u = x 2 first.
19. - - -4 + 4 log(l - x) 20. -4n I( I ) 1 4
4 I - x
ANSWERS TO EXERCISES A65
21. - Be - B - e - B + 1. Yes, 1. First evaluate the indefinite integral J xe - x dx by parts. Let
u = x, dv = e - x dx,
du = dx, v = _e- X •
Then the integral is equal to
Now put in the limits of integration:
As B -+ 00 the two terms with B approach 0, so you get the answer.
22. Yes, 5/e 23. Yes, 16/e
1 1 24. - log B + log 2· Yes, l/1og 2. Let u = log x. When x = 2, u = log 2. When
x = B, u = log B. Then
fB ----c- l--;;:2 dx = rOgBu -2 duo
2 x(logx) J1082
25. Yes, 1/3(1og W 26. 2 log 2 - 2. The indefinite integral is x log x-x. Let 0 < a < 2. Then
J.210g x dx = x log x - x I: = 2 log 2 - 2 - (a log a - a).
As a -+ 0 we know from Chapter VIII, §5, Exercise 14 that a log a approaches O. Hence a log a - a approaches 0 as a approaches 0, which gives the answer.
XI, §2, Supplementary Exercises, p. 346
1. !(x2 arctan x + arctan x - x). Use u = arctan x and dv = x dx. Also use the trick
f x 2 x~ 1 dx = f x 2 x~ ~ ~ 1 dx = f 1 dx - f x2 ~ 1 dx.
2. (a) If 1 is the integral, then
1 = x~ + arcsin x-I
so 21 = x~ + arcsin x, and dividing by 2 yields the answer.
A66 ANSWERS TO EXERCISES
(b) Integrating by parts reduces the integral to (a): u = arcsin x, dv = x dx. Use a trick as in Exercise 1.
~ 1t 1t 2 3. !(2X2 arccos x - arccos x - Xv' 1 - x 2) 5. 1 -"2 7. 32 9. ~
10.
Let u=l-x2, du=-2xdx. When x=O, u=1 and when x=l, u=O. Then
11. - 2e - fi(X3/2 + 3x + 6XI /2 + 6). Let x = u2 first. Then
13. Let u = (log x)" and dv = dx. 14. Let u = x" and dv = eX dx. 15. Let u = (log x)" and dv = xm dx. Then
1 du = n(log X)"-I - dx
x and v = xm+I/(m + I).
16. First we find the indefinite integral by parts with u = x", dv = e- X dx, so
Then we have the definite integral
Taking the limit as B -+ 00 and using that B"e- B -+ 0, we find :
ANSWERS TO EXERCISES A67
Let In = SO' xne- x dx. This last equality can be rewritten in the form
Thus we have reduced the evaluation of the integral to the next step. For instance, 1'0 = 1019; 19 = 918; 18 = 81 7 ; and so on. Continuing in this way, it takes n steps to get
_ , _, -x foo
In-n./o-n.oe dx,
where n! = n(n - 1 )(n - 2)· ··3 . 2 . 1 is the product of the first n integers. This final integral is easily evaluated, namely
foo e-X dx = lim fB e- X dx = lim _ e- X IB o 8-+00 0 8-00 0
= lim - [e- B - 1] = 1. B-oo
XI, §3, p. 354
1. -! sin3 x cos x - i sin x cos x + ix 2.! cos2 x sin x + ~ sin x
sin3 x sinS x . 2 3. - 3- - -5- 4. 3n 5. 8n 6. nab (If a, b > 0) 7. nr
8. (a) -2j2cos8/2 (b) 2j2sin8/2 13. -logcosx . x
14. arcsm 3
x h 1 bx 15. arcsin (; 16.! arcsin(y 2x) 17. - arcsin - Let x = au/b, dx = (a/b) duo
y 3 b a
18. (a) Co = an = 0 all n, bn = -(2/n) cos nn. (b) Co = n2/3, an = - (4/n 2 ) cos nn, b. = 0 all n. (c) Co = n/2, an = 2(cos mr - 1)/nn2 , bn = 0 all n.
19. (b) all an and Co = 0
XI, §3, Supplementary Exercises, p. 356
sin2 x 1. log sin x - --
2
2. Write tan2 x = tan2 x + 1 - 1 and note that d tan x/dx = tan2 x + 1.
n n n 3. -cos eX 4. Let x = 2u, dx = 2du 5. - 7. - 9 -
4 4 . 16
~ n . 1 11. t arcsin x - tx(l - 2x2)y 1 - x2 13. 2 14. -arcsm x - ~ JI=7 15. -16u'/2 + !U 3/2 where u = 16 - x2
A68 ANSWERS TO EXERCISES
16. Let u + 1 + x 2 • Then
The rest of the exercises are done by letting x = sin 0 or x = a sin 0, dx = a cos 0 dO. We give the answers, but work out Exercise 19 in full.
- I [a + J a2 - X2] a1 17. ~ log x 18. "2 arcsin(x/a) - !xJa2 - x 2
- J a2 - x2 1 [a + J a1 - X2] 19. 2 2 - -3 log . We have a choice of whether to let 2ax 2a x
x = a cos 0 or x = a sin 0. The principle is the same. Let us do as usual,
x = a sin 0, dx = a cos 0 dO.
Then
f--1-- dx = f 1 a cos OdO = ~ f_l_dO x 3 J a2 _ x 2 a3 sin 3 O(a cos 0) a3 sin3 0 .
It's a pain, but we show how to do it. Recall that to integrate positive powers of sine, we used integration by parts. We try a similar method here. Thus let
f l fl I f12 1= -=---30 dO = ~O ~O dO = -:--0 CSC 0 dO. sm sm sm sm
In analogy with the tangent, we have
so we let
Then
and so
whence
d cot 0 2 ----;,w- = -csc 0,
1 U = sin 0' dv = csc2 0 dO,
1 du = - ~O cos 0 dO,
sm v = -cot 0.
I = _ cot 0 _ f cos2 0 dO = _ cos 0 _ f 1 - sin 2 0 dO sin 0 sin 3 0 sin 1 0 sin 3 0
cosO f 1 I = - ~O - I + ~O dO, sm sm
I [ cos 0 ] I = 2 - sin2 0 - log(csc 0 + cot 0) .
ANSWERS TO EXERCISES A69
You may leave the answers in terms of (J, this is usually done. But if you want the answer in terms of x, then use:
x sin (J = -
a'
I a csc(J=--=
sin (J x' cos (J Ja 2 - x 2
cot (J = -.- = . sm (J x
I 20. - 2 cot (J where x = a sin (J, dx = a cos (J d(J.
a
~ (I+~) 21. v I - x 2 -log x . The method is the same as Exercise 19.
22. Let x = at, dx = a dt, and reduce to Exercise 14. x x
23. rr-::i - arcsin -v a2 - x 2 a
XI, §4, p. 370
1. -! log(x - I) + ¥ log (x + 7)
-I 2. 2(x2 _ 3)' Don't use partial fractions here, use the substitution u = x2 - 3
and du = 2x dx. 3. (a) H!og(x - 3) -log(x + 2)] (b) log(x + I) - log(x + 2) 4. - t log(x + 1) + 2 log(x + 2) - ~ log(x + 3)
I S. 2 log x -log(x + 1) 6. log(x + I) +-
x+1
2 7. -log(x + I) + log(x + 2) - -
x+2
x 8. log(x - 1) + log(x - 2) 9. 2 2 + t arctan x
(x + I)
I x 3 x 3 10. (a) 4 (x2 + 1)2 + 8 (x2 + I) + garctan x
II. X2-~ I - 3 [2(X2X + 1) + ~ arctan x]
I -I I x I x I x I x 12. "2 x2 + 9 + 18 x 2 + 9 + 54 arctan"3 13. 8 x 2 + 16 + 32 arctan 4
(x + 1)2 1 14. ~ log 2 + -2 arctan x. Factorization:
x + I
x 3 - I = (x - 1)(x2 + X + I) and X4 - I = (x + I)(x - 1)(x2 + I).
IS. C1 = -13to, C2 = - /do, C3 = -l~g, C4 = -lAg, Cs = /010
16. (a) Let x = bt, dx = bdt (b) Let x + a = bt, dx = bdt.
A70 ANSWERS TO EXERCISES
17. (a) -! arctan x + ! 10g(X - I) (b) Wog(x 2 - 1) - log(X2 + I)] x + 1
1 2x + 1 18. (a) ! log(x - I) - i log(X2 + X + I) - j3 arctan j3
1 x2 1 2x + 1 (b) "2log 2 1 - r:; arctan r:;
x+x+ y3 y3
19. -Iog(x - 1) + log(X2 + X + 1)
XI, §5, p. 377
1. 2~ - log(~ + 1) + log(~ - 1) 2. x -log(1 + eX)
3. arctan(eX ) 4. -Iog(~ + 1) + log(~ - 1) 5. Let y = f(x) = !(eX - eX) = sinh x. Then
But cosh x > 0 for all x, so f is strictly increasing for all x. If x is large negative, then eX is small, and e- X is large positive, so f(x) is large positive. If x is large positive, then eX is large positive, and e - X is small. Hence
f(x) -> 00
f(x) --+ - 00
as x -> 00,
as x -> - 00.
By the intermediate value theorem, the values of f(x) consist of all numbers. Hence the inverse function x = g(y) is defined for all numbers y. We have
1 1 1 1 g'(y) = - = - - = = ----===
f'(x) cosh x jsinh2 x + 1 .JY2+1.
ANSWERS TO EXERCISES A71
6. Let y = f(x) = hex + e- X) = cosh x . Then
rex) = !(eX - e- X) = sinh x .
If x > 0 then eX> 1 and 0 < e- X < 1 so rex) > 0 for all x> O. Hence f is strictly increasing, and the inverse function
x = g(y) = arccosh y
exists. We have f(O) = 1. As x -+ 00, eX -+ 00 and e- X -+ 0, so f(x) -+ 00.
Hence the values of f(x) consist of all numbers ~ 1 when x ~ O. Hence the inverse function g is defined for all numbers ~ 1. We have
1 1 1 1 g'(y)= - =--= =.
rex) sinh x Jcosh2 x-I R-=t
graph of f(x) = He' + e- Z ) = cosh x
Finally, let u = eX. Then y = !(u + l/u). Multiply by 2u and solve the quadratic equation to get
or eX = u = y - JY2=l. The graph of f(x) = cosh x for all numbers x bends as shown on the figure. and there are two possible inverse functions depending on whether we look at the interval
x;£O or x ~O.
Taking
x=log(y+~
is the inverse function for x ~ 0 and taking
x = log(y - R-=t) is the inverse functions for x ;£ O. Indeed suppose we take the solution with the minus sign. Then by simple algebra, you can see that
y-R-=t;£ 1.
An ANSWERS TO EXERCISES
[Proof: you have to check that y - 1 ~ JY2=l. Since y ~ 1, it suffices to check that (y - 1)2 ~ y2 - 1, which amounts to
or y ~ 1, which checks.]
Thus y - JY2=l ~ 1, whence
log(y - JY2=l) ~ O.
For x ~ 0 it follows that we have to use the solution of the quadratic equation for u in terms of y with the plus sign, that is
eX =u=y+JY2=l and x = log(y + JY2=l). 7. Let
Let x = 2 sinh t, dx = 2 cosh t dt. Then x 2 + 4 = 4 cosh2 t. Hence
8. log(x + ~) 9. Let
1= 2 cosh t dt f4 sinh2 t
2 cosh t
= 4 f sinh2 t dt = ~ f (e 2' - 2 + e- 2') dt
= te2' - 2t - te-2'.
1= dx. f X2 + I
x-~
Let x = sinh t, dx = cosh t dt. Then
f cosh2 t I = . h h cosh t dt sm t - cos t
ANSWERS TO EXERCISES A73
10. -t log(x + Jx2=l) + txJx2=l (see Exercise 11)
11. -t 10g(B + JB2=1) + tBJB2=1. The graph of the equation x 2 -y2 = 1 is a hyperbola as drawn.
" " grapb of
y=JT-i
,
The top part of the hyperbola in the first quadrant is the graph of the function
Y= Jx2=l,
with the positive square root, and x ~ 1. Hence the area under the graph between 1 and B is
Area = r Jx2=l dx.
We want to make the expression under the square into a perfect square. We use the substitution
x = cosh t and dx = sinh t dt.
Then you get into an integral consisting of powers of e' and e-' which is easy to evaluate. Change the limits of integration in the way explained in the last example of the section. Let u = e' and solve a quadratic equation for u. You will find the given answer.
12. 10g(B + JB2+l) + BJB2+l 13. Let y = a cosh(x/a). Then
dy 1 -d = a sinh(x/a) · - = sinh(x/a),
x a
d2y 1 1 - = cosh(x/a) · - = - cosh(x/a) dx 2 a a
1 = - Jl + sinh2(x/a)
a
1 = - Jl + (dy/dx)2.
a
14. Let x = az, dx = adz, and reduce to the worked-out case.
A74 ANSWERS TO EXERCISES
XII, § 1 p. 384
n2 n n2 n 2 . 54n 1. ~nr3 2. n 3. 8 - 4 4. 8 + 4 5. -3- 6. n(e - 2) 7. ne2
8. n[2(1og 2)2 - 4 log 2 + 2]. Integrate f (log X)2 dx by parts, u = (log X)2,
210gx du = -- dx, dv = dx
x
9. 16n 10. (a) ~ G2 - e!B} yes 2:2
(b) ~ (~- )B} yes 4:4 (c) ~ (~- )B} yes 4:2
r nr2h ll. Equation of the line is y = h x. Volume is -3- 12. 2n(1 - Ja), 2n
n n n 13. 24 - 383 ; yes 24 14. For all c> 1/2, n/(2c - I)
15. For all c < 1/2, n/(\ - 2c)
XII, §1, Supplementary Exercises, p. 385
1. f(x) = R + Ja 2 - x 2 and g(x) = R - Ja2 - x 2. The volume is
V = n ro f(X)2 dx - n rog(X)2 dx
which after easy algebra comes out equal to
4nR f:o J a2 - x 2 dx = 2n2 Ra2.
32n 2n 5n n 4 2 n - 2 - 10 2. - 5- 3. 12n 4. 2n 5. 3" 6. 14 7' "3 8. 3M b 9'"2 (e - e )
10. n[2(1og 2)2 - 410g 2 + 2] II. { J3 - 3J 12. n( 1 - ~). n as B -+ 00
13 ~ (I -~) ~ as 8 -+ 00 . 3 8 3 ' 3 14. n log 8
1 15. n log - . No limit as a -+ O. The volume increases without bound.
a
16. nG - I)- No limit as a -+ O. The volume increases without bound.
17. { l! -cos a + log(fi - I) -Iog(csc a - cot a)J
No limit as a -+ O. The volume increases without bound.
XII, §2, p. 390
I. 6n 2. a2 (using symmetry and values of 8 such that sin 28 ~ 0, the problem reduces to f0'2 a2 sin 28 d8)
ANSWERS TO EXERCISES
n 3. na 2 4. 12 5. 3n/2 6. 3n/2 7. 9n/2 8. n/3
XII, §2, Supplementary Exercises, p. 390
3n 9n 3n 3n 3)3 3n l. 25n 2."2 3. 1t 4. "2 5. 8 6."2 7. 2n + -2- 8. "2
n 9n 2 2 4 4 5}5 9. 4 10."2 11. 103 12. 103 13. 3 14. 3 15. - 6- 16. 10
XII, §3, p. 397
l. -/-r(103/2 - 1) 2. {j + lOge: 2$) 3. P+t + 2 - fi + IOg( I JA.)
1 + e4 + I (Ji7 + 4)114
4. 2Ji7 + log ~ 17 - 4
5. v I + e2 + ! log v 1 -r- e - V 2 + ! log --~ ~-1 ( ;::, fi-I) Jl+7+1 fi+l
I 6. 217(31 3/2 - 133/2) 7. e - -
e
8. We work out Exercise 8 in full .
f3/4J ( 2X)2 length = 0 1 + 1 _ x 2 dx
_ f3 /4 I + x2 - 1 2 dx
o - x
_ f3 /4 _2_ f3 /4 x2 - 1 - 1 2 dx + 1 2 dx
o -x o-X
= 2 - - - + -- dx - dx f3/4 I (I 1) f 3/4
o 2 I-x l+x 0
= -Iog(l - x) + log(1 + x) --1
3/4 3
o 4
( I + 3/4) = log 1 _ 3/4 - i = log 7 - 3/4
A75
A76 ANSWERS TO EXERCISES
9. ~ (e -n 10. log(2 +)3)
XII, §4, p. 407
2. 2nr 3. )2(e2 - e) 4. (a) i (b) 3 5. 2j5 + 10g(j5 + 2)
6. 4)2 + 2 log ~ + 1 7. 2)3 8. 5 9. 8 10. 4a 12. )2(e2 - e) ",2-1
!f7 3n 13. )2(e'2 - e") 14. (83/2 - 53/2) 15. -"'4-1
t (e- 4 - e- s) 16'"4
fo [8+2foJ 17. j5 - -4- + log 1 + j5 18. 4 sin i = 2)2 -)2 19. 4 20. 2
21. 8 22. n 23. 2)3
XII, §5, p. 415
I. 12na2/5 2. ;7 (lOjiO - I) 3. 2; (26)26 - 2)2) 4. 4n2a2 5. 4n2aR
6. ~ (17fo - I)
XII, §6, p. 418 10 180 [ n 1 n] 1. 5 Ib/in.; 80 in.-Ib 2. ~ Ib/in.; --;- cot 9 - 2 csc 9 in.-Ib
SIn 9
[1 1] C 99c. . . 3. c - - - 4. yes; - 5. 200 where C IS the constant of proportlOnahty
r1 r r 1
6. 2 x 106 pound-miles 7. ~ in.-Ib
8. (a) -90 CmM dyne-cm (b) 9 CmM dyne-em
10. 1500 log 2 in.-pounds. If A is the area of the cross section of the cylinder, and P(x) is the pressure, then
P(x)· Ax = C = constant.
If a is the length of the cylinder when the volume is 75 in3, then
Work = Force dx = P(x)A dx = - dx = Clog 2. f2. f2. f2. C
Q Q Q X
But C = 20·75 = 1,500 from initial data. This gives the answer.
ANSWERS TO EXERCISES A77
XII, §7, p. 423
2. 10 3. lOfJog 3
XIII, §1, p. 434
( _I)k+ l(k - 1)' 1. (a) f(k)(X) = (1 k •
+ x) (b) Pk)(O) = (-1)k+ l(k - 1)!
(c) Since (k - 1)!jk! = 11k, we get from (b)
Pk)(O) (-lrl(k-l)! (-lrl --------
k! k! k
and
This proves that when f(x) = 10g(1 + x) then the n-th Taylor polynomial is given by
2. For f(x) = cos x, pn)(x) = pn+4)(x). Use this and the formula for p.(x) to derive p.(x) for the function f(x) = cos x.
XIII, §3, p. 446
x 2 X4 1.1 - - + -
2! 4! 2. I j<·)(c) I ~ 1 for all n and all numbers c so the estimate follows from
Theorem 2.1. 0.01 (01)3 1
3. 1 - -2- + R4 (0.1) = 0.995 + R4 (0.1) 4. IR31 ~ ---i! = 6 X 10- 3.
5. IR41 ~ ~1O-4 x3
6. P4(X)=X+ 3 . Let f(x)=tanx . You have to find first all derivatives
f(l)(x), P2)(X), P3)(X), P4)(X), and then PI)(O), ... ,f(4)(0). Then use the general formula for the Taylor polynomial
7. IRsl ~ 10- 4 by crude estimates
A78 ANSWERS TO EXERCISES
8. (a) SinG + 1;0) = ~ + !! C;O) + E and lEI < 10- 3
(n n) J3 l(n) (b) COS - + - = - - - - + E 6 180 2 2 180
(n 2n ) fi fi ( 2n ) (c) sin -+ - =-+ - - +E 4 180 2 2 180
(n 2n ) fi fi (2n) (d) cos -+- =--- - +E 4 180 2 2 180
(n 2n ) 1 J3 (2n) (e) sin - + - = - + - - + E 6 180 2 2 180
(n 2n ) J3 1 (2n) (f) cos 6 + 180 = 2 - 2 180 + E
We carry out the steps for part (e) in full.
(n 2n ) (n n ) sin 32° = sin 6 + 180 = sin 6 + 90 = sin(a + h)
= sin - + cos - - + R (h) n ( n) n 6 6 90 2
and
R - :S - - :S - - :S (3.5 x 10- ) -:S 10- . I ( n ) I ( n)2 1 (3.15)2 1 2 2 I 3 290 -902- 902- 2-
(n n) J3 I(n) 9. cos 6 + 180 = 2 - 2 180 + E
. (n n) J3 I ( n ) 10. sm "3 + 180 = 2 + 2 180 + E
sin x x2 Rs(x) 12. (a) - = 1 - - + -- so
x 3! x
- dx=I- - +E fl sin x I
o x 3· 3! where
(0.1 )2 (b) --+E
4 where
10- 4
lEI ~ 4.4!
ANSWERS TO EXERCISES A79
U 3
(C) Write sin U = U - 3! + Rs(u), so that for u = x 2 we get
We have, for u ;;; 0,
Hence
fill o sin x 2 dx = "3 - 7 . 3 ! + E,
where
IEI:$ - dx = --:$ 10- 3. fl XIO 1
- 0 5! 11 . 5! -
and
1 1 (e) 1 - 5. 2! + 9. 4! + E,
1 and lEI ~ 13.6!
1 (f) I - 5 . 3! + E, and
1 lEI ~ 9.5!
13.
= fl/2 (_ ~ + x3 + R6(X») dx o 2 4! x
= - - +- +E x 2 x411/2 4 96 0
1 1 = - 16 + 16.96 + E,
and
XIII, §4, p. 448
A80 ANSWERS TO EXERCISES
10-' (10- 2)3 2(10- 6 ) 10- 6
4.IR I~e --~--=--3 - 3! - 6 3
_1 1-11' 1 1 6. IR71 ~ e ~ ~ 7! = 5040
7. (a) IR41 ~ e21!:4 ~ 6 (b) IR41 ~ e31 !:4 ~ 214
1215 64 IW 648 8. (a) IRsl ~ e2 ST ~ 15 (b) IRsl ~ e3 ST ~ 5
16 . 212 16.216 9. (a) IRI31 ~ ~ using e < 4 (b) IRI61 ~ l6! using e < 4
13 1 -2 12 (-2t 213
10. e = 1 + L ,+ E II. e = L -k-'- + E and lEI ~ -3' .= I n. k=O • 1 .
1 lIe 12. (a) 1 + - + - + ... + - + E, where lEI ~ -
2·2! 3 · 3! 7·7! -8 · 8!
u4
(b) Write e" = 1 + u + ... + , + Rs(u). For u ~ 0 we have e" ~ 1, and 4.
luiS hence IRs(u)1 ~ ST' Now put u = _x2• Then
X4 x 6 x 8
e - x' = 1 - x 2 + - - - + - + R (- x 2). 2! 3! 4! 5
Integrate the first part consisting of powers of x term by term. We get
fl -x2 1 I 1 1 ° e dx = 1 -:3 + 5. 2! - 7 . 3! + 9. 4! + E,
and
1 1 I 1 1 (c) 1 + :3 + 5 . 2! + 7 . 3! + 9 . 4! + 11 . 5! + E,
e 1 where lEI ~ 13 . 6! <:3 x 10- 3.
(d) e" = 1 + u + R2(u), and we put u = x 2. For 0 ~ x ~ 0.1 we find e" ~ 2 (generous estimate). Hence
2u2
IR 2(u)1 ~ 2! = u2 .
eX dx = 0.1 + -- + E, fo. 1 2 (0.1)3
° 3 where
(e) 0.1 - (!)1O- 3 + E, where lEI ~ 10- 6 .
ANSWERS TO EXERCISES
XIII, §5, p. 455
(0.2)2 (0.2)3 _ 4 1. (a) log 1.2 = 0.2 - -2- + - 3- + R4 , IR41 ~ 4 · 10
(b) log 0.9 = -log 10/9 = -IOg(1 +~)
[ I (1/9)2J I I = - 9 - - 2- - R3(1 /9) = - 9 + 2 . 81 - Ril/9)
and I-R3(l /9)1 <! x 10- 3
(O.oW 53 (c) log 1.05 = 0.05 - -2- + R3 and IR31 ~ "3 .10- 6
(d) log 9/ 10 = -log 10/9 as in (b).
(e) log 24/25 = -log 25/24 = -IOg(1 +214)
-I I = 24 + 2(24)2 - R 3(l /24),
and IR3(1/24)1 < 10- 3. (f) log 26/25 = 0.04 + Ril/25), IR21 ~ 8 x 10- 4
A81
Z. (a) We transform the right-hand side until it is equal to the left-hand side. We have :
-710g ~ + 210/4 + 3 log 81 = IOg(1O)7(24)2(81)3 10 25 80 9 25 80
2757(23 . W(34)3
= log (32)'54(24)353
27263231257 = log 21231457
= log 2.
The case of log 3 is done in the same way. Note that each fraction 9/ 10, 24/25, 81 /80 is close to 1, and hence the value of the log is approximated very well by just a few terms from the Taylor formula. For instance,
Then
and
log - = -log - = -log I + -9 10 (I) 10 9 9 .
( I) I (I/W (1/9)3 (I/W (1/9)5 9 log 1+ - =-- - -+ - ---- +-- +R (1/ )
992345 6 '
A82 ANSWERS TO EXERCISES
Hence
(I)
where E) = 7R6(1 /9) and
Observe the factor of 7 which comes in throughout at this point. Next, we have
and
Hence
(2)
where
25 ( 1 ) 1 1 1 log 24 = log 1 + 24 = 24 - 242 + 243 + R4(1 /24)
= A2 + R4(1/24),
25 -2log - = -2A2 - 2R4(1 /24) = -2A2 + E2 ,
24
Thirdly, we have
81 ( 1 ) 1 1 log 80 = log I + 80 = 80 - 802 + R3(1 /80)
= A3 + R3(1 /80),
and
Hence
(3)
where
We may now put together the computations of the three terms, and we find:
log 2 = 7A) - 2A2 + 3A3 + E,
and
ANSWERS TO EXERCISES
lEI ~ IE11 + IE21 + IE31 < 3.5 x 10- 6 + 2 X 10- 6 + 2 X 10- 6
< 10- 5 .
This concludes the computation of log 2.
A83
The computation for log 3 is similar. In each case, note that the factors 7, - 2, 3 for log 2 and 11, - 3, 5 for log 3 have to be taken into account, and contribute to the error term.
XIII, §6, p. 459
1. u + u 2u
2 arctan u = arctan u + arctan u = arctan - - 2 = arctan - - 2 ' l-u l-u
2u 3 arctan u = arctan u + 2 arctan u = arctan u + arctan - -2 .
1 - u
Now let v = 2u/(l - u2) and use the formula for arctan u + arctan v.
2. (a) Let u = 1/2 and v = 1/3 in the formula for arctan u + arctan v. (b) Let u = 1/5 and v = 1/8 in this same formula. (c) We have to apply the addition formula repeatedly. We start with
Next,
Next,
1 1 1 1/5 + 1/5 2 arctan 5 = arctan 5 + arctan 5 = arctan 1 _ 1/25
5 = arctan -
12 '
5 1 5/ 12 + 1/7 47 arctan 12 + arctan "1 = arctan 1 _ 5/84 = arctan 79 '
1 1 1 1/8 + 1/8 2 arctan "8 = arctan "8 + arctan "8 = arctan 1 _ 1/64
16 = arctan 63 '
Then finally
1 1 1 47 16 2 arctan 5 + arctan "1 + 2 arctan 8 = arctan 79 + arctan 63
47/79 + 16/63 = arctan - --'----- --'----
1 - 47· 16/79 · 63
= arctan 1 = rt/4.
A84 ANSWERS TO EXERCISES
3. By Taylor's formula, we have
so
(1)
where
and IE I::; 8 - - = - x 10-1 (1)5 16 4
I - 5 5 3 .
Second, we have
so
(2) 4 . arctan - = - - - - + E 1 4 4 (1)3 7 7 3 7 2'
where
and IE I ::; 4 - - < - X 10- 4 . 1 (1)5 2 2 - 5 7 5
Third, we have
so
(3) 8 . arctan - = 1 - - - + E 1 8 (1)3 8 3 8 3
where
and
Adding the three expressions, we find by Exercise 2(c):
1 1 1 1t = 8 arctan 5 + 4 arctan "1 + 8 arctan 8
where
and
ANSWERS TO EXERCISES
4. arctan(I /5) + arctan(l /5) = arctan 5/ 12 using u = v = 1/5, arctan(I /5) + arctan(5/ 12) = arctan(37/55) using u = 1/5 and v = 5/12, arctan(I /5) + arctan(37/55) = arctan(120/119).
This last value is equal to 4 arctan(I/5). Then
4 arctan(l/ 5) - arctan(I /239) = arctan(120/ 119) + arctan( -1/239).
Let u = 120/ 119 and v = -1/239 and use arithmetic to get arctan 1.
XIII, §7, p. 467
A8S
In the answers, we give only the approximating value, except in a couple of cases to illustrate an estimate for the error term. But you should include the estimate in your work.
3 1. (a) IR21 ~ 32.10-4 < 10- 5• We use s = 1/4 and
11 31 1 7/4 2 Rz{x) = 4 - 4 2: (1 + c)- Ixl·
Since (l + c) -7/4 ~ 1, we get
3 3 IR 2(0.1)1 ~ 32 (0.1)2 ~ 32 10-4 < 10- 5•
3 3 (b) IR21 ~ 8. 10- 2 (c) IR21 ~ 32 . 10- 2
1 1 2. (a) IR31 ~ 5.10-4 (b) IR31 ~ 16 (0.8)-5/ 2(0.2)3 ~ 2. 83 ~ 10- 3
1 (c) IR31 ~ 16.10-4
3. Estimate R2(x) for (1 + x)! /3 and -0.1 ~ x ~ 0.1 . The general expression for R2 with s = 1/3 is
so the term (1 + C)-5 f3 = 1/(1 + C)5 /3 will be biggest when x = -0.1. Also Ixl2 is biggest when x = 0.1. Hence
~! (10)5/310 _2
- 9 9
<! (10)2 10 _2 = _1_ <! x 10-2. = 9 9 729 7
A86 ANSWERS TO EXERCISES
1 4. (a) IR 1:0;1.(0.8)-3/2.10- 2 :0;1.--.10- 2 :0;10- 2 (b) IR 1:0;1.10- 2
2 - 2 - 2 (0.8)2 - 2 - 8
5. (a) 5(1 + ~'1~5) + E (b) 5(1 +~ . 215 -~. 6~5) + E
( 1 6 1 62 ) (c) 5 1 + 3' 125 - 9' 1252 + E.
In this part we include the estimate for the error. We write
131 = 125 + 6 = 125(1 + 1~5) so
( 6 )1/3
(131)1/3 = 5 1 + 125 .
Then
1 ( 6 ) 1 1 2 5 1 ( 6)3 1 4 R3 125 ~ 3 3 3 3! 125 ~ 9 x 10- .
Hence
131 - 5 1 - - ---1/3 ( 1 6 1 62)
( ) - + 3 125 - 9 (12W + E,
where
(d) 6(1 + ~. :3) + E
6 (a) 1O(l-!'~-!'~)+E (b) 1O(l+!'~)+E . 2 100 8 1002 2 100
(c) 10(1 +!'~-!'~)+E 2 100 8 1002
( 1 3 1 32 1 3 33 ) (d) 5 1 + 2' 25 - 8' 252 + 3!' 8' 253 + E
By writing 28 = 25 + 3 = 25(1 + 3/25) you can apply the same method as in the examples, and E = 5Ri3/25), so we have to estimate R4(3/25). We have:
1 ( 3)111351(3)4 1 4 R4 25 ~ 2 2: 2: 2: 4! 25 ~ 8 x 10-
so lEI ~ (5/8) x 10- 4 which is within the desired accuracy.
XIII, §8, p. 471
1. 0 2. 1/4! 3. 2 4. 0 5. 1 6. 1 7. 1 8. 1 9. 1 10. 2 11.! 12. 0 13. -! 14. 1 15. 1 16. 1 17. -! 18. -1 19. 2 20.! 21. 1 22. 1
ANSWERS TO EXERCISES A8?
23. -1 24. 1 25. I 26. 0 27. -i 28. 0 29. t 30. 1 31. -t 32.-~
1 1 33. (a) 0 (b) 0 (c) 0 34. 1 35. -2 36. 0 37., 38. 0 39. -1 5.
XIV, §2, p. 480
3. No 4. Yes 5. No 6. No 7. No 8. Yes 9. Yes
XIV, §3, p. 482
1. Yes 2. Yes 3. No 4. Yes 5. No 6. Yes 7. No 8. Yes 9. No 10. Yes 11. No 12. Yes 13. No 14. Yes 15. Yes 16. Yes 17. Yes 18. Yes
XIV, §4, p. 485
3. Yes 4. Yes 5. Yes 6. Yes 7. Yes 8. Yes 9. Yes 10. Yes
XIV, §5, p. 488
1. Yes 2. Yes 3. Yes 4. Yes 5. Yes 6. Converges, but not absolutely 7. Yes 8. Converges, but not absolutely 9. Converges, but not absolutely 11. Converges, but not absolutely
12. Converges, but not absolutely 13. Does not converge; does not converge absolutely 14. Does not converge 15. Converges, but not absolutely 17. Converges, but not absolutely 18. Yes 19. Converges, but not absolutely 20. Converges, but not absolutely
XIV, §6, p. 494
2. (a) 4je2 (b) 2255e- 4 j3 3 3. (a) 0 (b) 00 4. 1 5. 1 6. 1 7. 1 8. t 1 1 1 4
9. 2 10. 0 11. 1 12. 1 13. 2 14. 1 15. 4 16. ~ 17. 27 18. 2" 19. 0 e e
20. 2 21. 2 22. 3 23. 1 24. 00 25. 1 26. 00 27. 1 28. 00 29. e 30. 00
App., §1, p. 504
1. (a) glb is 2; lub does not exist. (b) glb is 1; lub does not exist. (c) glb does not exist; lub does not exist.
2. (a) glb is 0; lub is 0. (b) glb is 0; lub is 0. (c) glb is -2; lub is 2. (d) glb does not exist; lub is ¥.
A88 ANSWERS TO EXERCISES
App., §2, p. 513
4. f(x) exist for all x; f(x) = 1, Ixl ~ 1; f(x) = 0, Ixl < 1 5. (a) f(l) = 0, fm = -1, f(2) = 1
(b) lim f(x) does no exist (c) lim f(x) does not exist x--l
6. (a) f(1) = 0, f(!) = 1, f(2) = 1 (b) lim f(x) = 1 (c) lim f(x) = 1
x--t
7. (a) 0 (b) 0 (c) 0 (d) 0 8. (a) 0 (b) 0 (c) 0
XV, §1, p. 530
A+B A-B 3A -2B
1. (1,0) (3, -2) (6, -3) (2, -2) 2. (-1,7) (-1, -1) (-3,9) (0, -8) 3. (1,0,6) (3, -2,4) (6, -3,15) (2, -2, -2) 4. (-2,1, -1) (0, -5,7) (-3, -6,9) (2, -6,8) 5. (3n, 0, 6) (-n,6, -8) (3n,9, - 3) (-4n,6, -14) 6. (15 + n, 1,3) (15 - n, -5,5) (45, -6, 12) (-2n, -6,2)
XV, §2, p. 534
1. No 2. Yes 3. No 4. Yes 5. No 6. Yes 7. Yes 8. No
XV, §3, p. 537
1. (a) 5 (b) 10 (c) 30 (d) 14 (e) n2 + 10 (f) 245 2. (a) -3 (b) 12 (c) 2 (d) -17 (e) 2n2 - 16 (0 15n - 10 4. (b) and (d)
XV, §4, p. 551
1. (a) J5 (b) jiO (c) J30 (d) Ji4 (e) ~ (f) J24s 2. (a) j2 (b) 4 (c) j3 (d) J26 (e) )58 + 4n2 (f»)10 + n2
3. (a) (~, -~) (b) (0, 3) (c) (-i, L i) (d) (!t, -ii, M) n2 - 8 15n - 10
(e) 2n2 + 29 (2Jr, - 3, 7) (f) 10 + n2 (n, 3, -1)
4. (a) (-~,~) (b) (-~, ¥) (c) (is, - /5'!) (d) -ffi -1, -2, 3)
2n2 - 16 3n - 2 (e) n2 + 10 (n, 3, -1) (f) ~ (15, -2,4)
ANSWERS TO EXERCISES
5. (a) J5-fi (b) -2 (c) 10 (d) 13 (e) 2 5 34 J5 ji4 J35 j2l Jil ji2 35 6 1 16 25
6. (a) ~' {;t1£'0 (b) ~' ~' M£A1 ",,41·35 ",,41·6 ",,17·26 ",,41·17 ",,26·41
7. Let us dot the sum
with Ai. We find
Since A j • Ai = 0 if j "# i we find
But Ai· Ai "# 0 by assumption. Hence Ci = 0, as was to be shown. 8. (a) IIA + BI12 + IIA - BII2 = (A + B)·(A + B) + (A - B)·(A - B)
= A 2 + 2A . B + B2 + A 2 - 2A . B + B2
= 2A2 + 2B2 = 211AI12 + 211BII2
9. IIA - BI12 = A2 - 2A·B + B2 = IIAII2 - 211AII IIBllcos () + IIBII2
XV, §5, p. 556
A89
1. (a) Let A = P2 - PI = (-5, -2,3). Parametric representation of the line is X(t) = PI +tA=(I, 3, -1)+t(-5, -2,3). (b) (-1, 5, 3)+t(-I, -1,4)
2. X = (1, 1, -1) + t(3, 0, -4) 3. X = (- 1, 5, 2) + t( -4, 9, 1) 4. (a) (-~, 4,!) (b) (-1, ¥, 0), (-i, ¥, 1) (c) (0, 157 , -%) (d) (-1,1/-,!)
P+Q 5. P + !(Q - P)=-2-
XV, §6, p. 562
1. The normal vectors (2, 3) and (5, - 5) are not perpendicular because their dot product 10 - 15 = -5 is not O.
2. The normal vectors are (-m, 1) and (-m', 1), and their dot product is mm' + 1. The vectors are perpendicular if and only if this dot product is 0, which is equivalent with mm' = -1.
3. Y = x + 8 4. 4y = 5x - 7 6. (c) and (d) 7. (a) x-y+3z= -1 (b) 3x+2y-4z=21t+26 (c) x-5z= -33 8. (a) 2x + y + 2z = 7 (b) 7x - 8y - 9z = -29 (c) y + z = 1 9. (3, -9, -5), (1,5, -7) (Others would be constant multiples of these.)
10. ( - 2, 1, 5) 11. (11, 13, -7)
A90 ANSWERS TO EXERCISES
12. (a) X=(I , 0, -1)+t( - 2, 1, 5) (b) X = (-10, -13, 7) + t(11 , 13, -7) or also (1, 0, 0) + t(11, 13, -7)
1 2 4 2 13. (a) -3 (b) - r;;:, (c) !££ (d) - f10
y'42 y'66 y'18
14. (a) (-4, ¥, ¥) (b) (H, M, -lJ) 15. (1, 3, - 2)
8 13 16. (a) r:;c (b) M1
y'35 y'21
17. (a) -2/)40 (b) (41/ 17, 23/ 17) 18. (a) x + 2y = 3 (c) 6/)5
19. -12/7j6
XVI, §1, p. 575
1. (e', -sint,cost) 2. (2 cos 2t,_I _ , 1) 3. (-sint,cost) 1 + t
4. (-3 sin 3t, 3 cos 3t) 6. B
7. G' f) + {~ , f} (-1, 0) + t( -1, 0), or y = j3x, y = 0
8. (a) ex + y + 2z = e2 + 3 (b) x + y = 1
11. J(X(t) - Q). (X(t) - Q)
If to is a value of t which minimizes the distance, then it also minimizes the square of the distance, which is easier to work with because it does not involve the square root sign. Let f(t) be the square of the distance, so
f(t) = (X(t) - Q)2 = (X(t) - Q) . (X(t) - Q).
At a minimum, the derivative must be 0, and the derivative is
f'(t) = 2(X(t) - Q) . X'(t).
Hence at a minimum, we have (X(to) - Q) . X'(to) = 0, and hence X(to) - Q is perpendicular to X '(to), i.e. is perpendicular to the curve. If X(t) = P + tA is the parametric representation of a line, then X '(t) = A, so we find
(P + toA - Q). A = O.
Solving for to yields (P - Q) . A + toA . A = 0, whence
(Q - P)· A to = .
A · A
13. Differentiate X'(t) 2 = constant to get
2X'(t)· X"(t) = O.
ANSWERS TO EXERCISES A91
14. Let v(t) = IIX'(t)ll. To show v(t) is constant, it suffices to prove that v(t)2 is constant, and V(t)2 = X'(t)· X'(t). To show that a function is constant it suffices to prove that its derivative is 0, and we have
d dt V(t)2 = 2X'(t) . X"(t).
By assumption, X'(t) is perpendicular to X"(t), so the right-hand side is 0, as desired.
15. Differentiate the relation X(t)· B = t, you get
X '(t)· B = 1,
so IIX'(t)IIIIBII cos {} = 1. Hence IIX'(t)11 = 1/ IIBII cos {} is constant. Hence the square X'(t)2 is constant. Differentiate, you get
2X'(t) · X"(t) = 0,
so X'(t)· X"(t) = 0, and X'(t) is perpendicular to X"(t), as desired.
16. (a) (0, 1, n/8) + t( -4, 0, 1) (b) (1, 2, 1) + t(1, 2, 2)
(c) (e\ e- J , 3J2) + t(3e- J , -3e-\ 3J2) (d) (1, 1, 1) + t(l, 3,4)
18. Let X(t) = (e', e2', 1 - e-') and Y({}) = (1 - {}, cos {}, sin (}). Then the two curves intersect when t = 0 and {} = O. Also
and Y'({}) = (-1, -sin {}, cos (}) so
X'(O) = (1, 2, 1) and Y'(O) = (-1, 0, 1).
The angle between their tangents at the point of intersection is the angle between X'(O) and Y'(O), which is n/2, because
. X'(O) · Y'(O) cosme of the angle = II X'(O) II II Y'(O)II = o.
19. (18, 4, 12) when t = - 3 and (2, 0, 4) when t = 1. By definition, a point X(t) = (x(t), y(t), z(t» lies on the plane if and only if
3x(t) - 14y(t) + z(t) - 10 = O.
In the present case, this means that
This is a quadratic equation for t, which you solve by the quadratic formula. You will get the two values t = - 3 or t = 1, which you substitute back in the parametric curve (2t 2, 1 - t, 3 + t2) to get the two points.
20. (a) Each coordinate of X(t) has derivative equal to 0, so each coordinate is constant, so X(t) = A for some constant A.
(b) X(t) = tA + B for constant vectors A#-O and B.
A92 ANSWERS TO EXERCISES
21. Let E = (0,0,1) be the unit vector 10 the direction of the z-axis. Then X'(t) = (- a sin t, a cos t, b) and
X'(t) · E b cos O(t) = II X'(t) II = J a2 + b2'
23. Differentiate the relation X(t)· B = e2', you get
X'(t) · B = 2e2' = IIX'(t)IIIIBII cos O.
But IIBII = I by assumption, so the speed is v(t) = IIX'(t)11 = 2e2'/cos O. Square this and differentiate. You find
8e4' X '(t) · X"(t) = --.
cos2 0
25. (a) To say that B(t) lies on the surface means that the coordinates of B(t) satisfy the equation of the surface, that is
Differentiate. You get
2z(t)z'(t) = 2x(t)x'(t) - 2y(t)y'(t),
which after dividing by 2 yields
Now
z(t)z'(t) = x(t)x'(t) - y(t)y'(t).
B(t)· B'(t) = x(t)x'(t) + y(t)y'(t) + z(t)z'(t).
= 2x(t)x'(t) by (*).
(b) Given any point (x ,y, z) the distance of this point to the yz-plane is just Ixl. So if x is positive, the distance is x itself. We use the derivative test: if x'(t) ~ 0 for all t then x is increasing. We have :
2x(t)x'(t) = B(t) · B'(t) by (a)
= IIB(t)IIIIB'(t)111 cos O(t).
By assumption, cos O(t) is positive, and the norms IIB(t)ll, IIB'(t)11 are ~ 0, so if x(t) > 0, dividing by 2x(t) shows that x'(t) ~ 0, whence x(t) is increasing, as was to be shown.
26. (a) (1 , 1, ~) + t(1, 2, 2) (b) x + 2y + 2z = 1
27. We have C(t) = (-e' sin t + e' cos t, e' cos t + e' sin t). Let 0 be the angle between C(t) and C(t) (the position vector). Then
O C(t)· C(t) cos =-----
lIC(t)IIIIC(t)11
and a little algebra will show you it is independent of t.
ANSWERS TO EXERCISES A93
XVI, §2, p. 579
1. j2 2. (a) 2JO (b) i jU
3 5( 6+ fo) 1 3. (a) 2 (fo - I) + 4 log 5 (b) e - ;
4. (a) 8 (b) 4 - 2j2
The integral for the length is L(t) = r J2 - 2 cos t dt. Use the formula
. 2 1 - cos 2u sm u = 2 '
with t = 2u.
2 + 2Y2 1 (Vs - 1 Y2 + I) 5. (a)Vs - Y2 + log Vs = Vs - Y2 + -2 log Vs ~ r:
1+ 5 5+1 v2-1
The speed is IIX'(t)1I = JI + (lft)2 so the length is
L = f2 ! j1+t2 dt = f~-/- du 1 t J2 U - 1
f ,/5 U2 - 1 + 1 f/5 f,/5 1 = 2 1 du = du + -2-1 duo
fi u- ,fi fiu-
But
These last integrals give you logs, with appropriate numbers in front.
(b) j26 - Jlo + -21 10g(:' - 1 . ~ + 1) = j26 - Jlo + log ~3 (I + :.) 26 + 1 10 - 1 1 + 26
6. 10g(j2 + 1) 7. 5/3 8. 8
XVII, §l, p. 586
I. 2.
Ellipses Parabolas
A94 ANSWERS TO EXERCISES
4.
Parabolas Hyperbolas
10. 11.
Lines Circles
12. c <0
c>o
c<o
Hyperbolas
ANSWERS TO EXERCISES A95
XVII, §2, p. 592
of/ax of/oy of/oz
1. y x 1
2. 2xi 5x2 y4 0 3. y cos(xy) x cos(xy) -sin(z) 4. - y sin(xy) -x sin(xy) 0 5. yz cos(xyz) xz cos(xyz) xy cos(xyz) 6. yzeXY ' xzeXY:Z xyeXY '
7. 2x sin(yz) x 2z cos(yz) x 2 y cos(yz) 8. yz xz xy 9. z+y z +x x+y
y x 10. cos(y - 3z) + -x sin(y - 3z) + ----= 3x sin(y - 3z)
J l - x 2i Jl- x 2i
11. (l) (2, 1, 1) (2) (64, 80, 0) (6) (6e6 , 3e6 , 2e6 ) (8) (6, 3, 2) (9) (5, 4, 3) 12. (4) (0, 0, 0) (5) (n 2 cos n2 , n cos n2 , n cos n2 )
(7) (2 sin n2 , n cos n2 , rr cos n2 )
13. (- 1, - 2, I) oxY oxY
14. - = YXy-1 - = xYlogx ax oy
XVII, §3, p. 598
1. 2, - 3 2. a, b 3. Q, b, c 5. Select first H = (h,O) = hE I . Then A· H = hal if A = (ai' Q2)' Divide both
sides of the relation
f(X + H) - f(X) = alh + Ihlg(H)
by h oj:. 0 and take the limit to see that a) = Dd(X) . Similarly use H = (0, h) = hE2 to see that a2 = Dd(x,y) . Similar argument for three variables.
XVIII, §l, p. 603
d 1. di (P + tA) = A, so this follows directly from the chain rule.
2. 5. Indeed, C'(t) = (2t, -3t- 4 , 1) and C(l) = (2, -3, I). Dot this with given gradf(1, 1, 1) to find 5.
A96 ANSWERS TO EXERCISES
3. C'(O) = (0, 1)
Let C'(O) = (a, b). Now grad f( C(O» = (9,2) and grad g( C(O» = (4, 1), so using the chain rule on the functions f and g, respectively, we obtain
2 = -dd f(C(t» I = (9,2)· (a, b) = 9a + 2b, t 1=0
1 = -dd g(C(t» I = (4, 1)· (a, b) = 4a + b. t 1=0
Solving for the above simultaneous equations yields C(O) = (0, 1). 4. (a) grad f(tP)· P.
(b) Use 4(a) and let t = O. 5. Viewing x, y as constant, put P = (x, y) and use Exercise 4(a). Then put
t = 1. If you expand out, you will find the stated answer.
7. (a) of/ax = x/r and of/oy = y/r if r = Jx2 + y2.
ij X ij Y ij (b) ;- = (2 2 2)1/2' ;- = (2 2 2)1/2' ;- = guess what? ux x + y + z uy X + y + z uz
or x · 8.-=~
ax; r
9. (a) of/ax = (3x 2y + 4x)COS(X3y + 2x2)
of/oy = x3 cos(x 3y + 2x2)
(b) of/ax = -(6xy - 4) sin(3x2y - 4x)
of/oy = -3x2 sin(3x2y - 4x)
2xy oj x 2 + 5 (c) oj/ax = - = -,,---
(x 2 y + 5y)' oy x 2y + 5y y
(d) oj/ax = !(2xy + 4)(x2 y + 4X)-1/2
oj/oy = !X2(X 2y + 4X)-1/2
XVIII, §2, p. 609
1.
(a) (b) (c) (d) (e) (f)
Plane
6x + 2y + 3z = 49 x+y+2z=2 13x + 15y + z = -15 6x - 2y + 15z = 22 4x + y + z = 13 z=O
Line
x = (6, 2, 3) + t(12, 4, 6) X = (1, 1, 0) + t(l, 1, 2) X = (2, -3, 4) + t(13, 15, 1) X = (1,7,2) + t(-6, 2, -15) X = (2, 1, 4) + t(8, 2, 2) X = (1, n/2, 0) + teO, 0, n/2 + 1)
2. (a) (3,0,1) (b) X = (log 3, 32n, -3) + t(3,0, 1)
(c) 3x + z = 3 log 3 - 3 3. (a) X = (3,2, -6) + t(2, -3,0) (b) X = (2, 1, -2) + t( -5,4, -3)
(c) X = (3, 2, 2) + t(2, 3, 0)
ANSWERS TO EXERCISES
4. IIC(t) - QII and see Exercise 11 of Chapter II, §l. S. (a) 6x + 8y - z = 25 (b) 16x + 12y - 125z = -75
(c) 1tX + y + z = 21t 6. x - 2y + z = 1 7. (b) x + y + 2z = 2 8. 3x - y + 6z = 14 9. (cos 3)x + (cos 3)y - z = 3 cos 3 - sin 3.
1 10. 3x + 5y + 4z = 18 II. (a) r,y; (5, 1, 1) (b) 5x + y + z - 6 = 0
...;27 -10
12. r,;:; 13. (a) 0 (b) 6 14. 4ex + 4ey + 4ez = 12e 3...; 12
XVIII, §3, p. 614
l. (a) i (b) max = JiO, min = -JiO 3 48 I1iC 2. (a) /C (b) TI (c) 2...; 145
2...;5
3. Increasing ( - 9f, -3f). decreasing C~, 3f) 4. (a) (2.:7/4' 2.~7/4' - 2 .:714) (b) (1, 2, -1,1)
S. (a) -2/fi (b) jli6 6. fi (5, 2, 5), 6}6 7. (fi' ~). fi 1
8. j3(2e - 5) 9. (a) 0 (b) -~
A97
10. For any unit vector A, the function of t given by f(P + tA) has a maximum at t = 0 (for small values of t), and hence its derivative is 0 at t = O. But its derivative is grad f(P + tA)· A, which at t = 0 is grad f(P)· A. This is true for all A, whence grad f(P) = o. (For instance, let A be anyone of the standard unit vectors in the directions of the coordinate axes.)
Although the above argument is the one which will work in Problem 11, there is a basically easier way to see the assertion. Fix all but one variable, and say Xl is the variable. Let
g(x) = f(x, a2, ... ,an), where P = (a l , · ·· ,an)·
Then 9 is a function of one variable, which has a maximum at X = a l .
Hence g'(a 1) = 0 by last year's calculus. But
g'(a l ) = Dtf(a l , ·· · ,an)·
Similarly DJ(P) = 0 for all i, as asserted.
XVIII, §4, p. 619
of dg or dg x l. - = - - = - - . Replace x by y and z. Square each term and add. You ox dr ox dr r
can factor
A98 ANSWERS TO EXERCISES
2. (a) -X/r3 (b) 2X (c) -3X/rs (d) -2e- r'X (e) -X/r2 (f) -4mX/r,"+2 (g) -(sin r)Xlr
3. F(/)2 = (cos t)2 A 2 + 2(cos t)(sin I) A . B + (sin 1)2B2 = 1, because A 2 = 8 2 = 1 since A, 8 are unit vectors and A· B = 0 by assumption. Hence IIF(t)11 = I, so F(t) lies on the sphere of radius l.
____ ----..F(t) = (cos t)A + (sin t)B
4. Note that L(t) = (1 - t)P + tQ. If L(t) = 0 for some value of t, then
(l - I)P = -IQ
Square both sides, use p2 = Q2 = 1 to get (1 - t)2 = t2. It follows that 1 -1
t = 1/2, so "i P = 2 Q, whence P = -Q.
5. By Exercise 4, L(t) =f. 0 if 0 ~ t ~ l. Then L(t)/ IIL(t)11 is a unit vector, and this expression is composed of differentiable expressions so is differentiable. Furthermore, we have
L(O) = P and L(l)=Q.
Thus if we put C(t) = L(t)/II L(t)ll, then IIC(t)1I = 1 for all t, and the curve C( t) lies on the sphere. Also
C(O) = P and C(I) = Q.
Hence C(t) is a curve on the sphere which joins P and Q. The picture looks as follows.
---:~Q ,-- -1-----, o
On the sphere cross section
Note that C(t) is the unit vector in the direction of L(t).
ANSWERS TO EXERCISES A99
6. Suppose P, Q are two points on the sphere, but P = - Q. In this case we cannot apply Exercise 5, but we can apply Exercise 3. We let
C(t) = (cos t)P + (sin t)A,
where A is a unit vector perpendicular to P. Then C(t)2 = 1, so C(t) lies on the sphere, and we have
C(O) = P, C(n) = -Po
Thus C(t) is a curve on the sphere joining P and -Po
7. Let x = a cos t and y = b sin t. 9. Let P, Q be two points on the sphere of radius a. It suffices to prove that
f(P) = f(Q). By Exercises 5 and 6, there exists a curve C(t) on the sphere which joins P and Q, that is C(t) is defined on an interval, and there are two numbers tl and t2 such that C(t 1) = P and C(t2) = Q. In those exercises, we did it only for the sphere or radius 1, but you can do it for a sphere of arbitrary radius a by considering aC(t) instead of the C(t) in Exercises 5 or 6. Now, it suffices to prove that the function f( C(t» is constant (as function of t). Take its derivative, get by the chain rule
d dt f( C(t» = grad f( C(t»· C'(t) = h( C(t»C(t)· C(t).
But C(t)2 = a2 because C(t) is on the sphere of radius a. Differentiating this with respect to t yields 2C(t)· C'(t) = 0, so C(t) · C'(t) = 0, which you plug in above to see that the derivative of f( C(t» = O. Hence f( C(t I»~ = f( C(t 2» so f(P) =f(Q)·
10. gradf(X) = (g'(r) ~,g'(r) ~,g'(r):) = g'(r) X (say in three variables), and r r r r
g'(r)j r is a scalar factor of X, so grad f(X) and X are parallel.
XVIII, §5, p. 623
k 1. k log /lXII 2. - 2r2 {
IOgr,
3. 1
(2 - k)r" 2'
k=2
Exercises 1 and 2 are special cases of 3. Let.
1 F(X) =;;: X.
We have to find a function g(r) such that if we put f(X) = g(r) then F(X) = gradf(X). This means we must solve the equation
1 g'(r) - X=-X rk r'
AlOO ANSWERS TO EXERCISES
or in other words
Then
g(r) = f r 1 - k dr,
which is an integral in one variable. You should know how to find it.
Index
A
Absolute value 8, 15 Absolutely convergent 486 Acceleration 107, 573 Accuracy 439 Addition formula 137, 457 d'Alembert 525 Alternating convergence 487 Angle between vectors 546 Angle of incidence 207 Application of integration 379 Applied maxima and minima 202 Approach 504 Arc 123 Arccosh 374 Arccosine 229 Arcsine 225 Arcsinh 374 Arctangent 229, 457 Area 118, 292, 323, 387, 411 Area in polar coordinates 388 Axes 21
B
Ball 502 Base of logarithm 259 Beginning point 531 Bending down 188 Bending up 188 Binomial coefficient 465 Binomial expansion 459 Bounded 502
C
Center of gravity 421 Chain rule 94, 600 Circle 37, 117 Circumference of circle 122 Closed ball and disc 541 Closed interval II Completing the square 39 Component 546 Composite function 93 Compound interest 260 Conservation law 622 Constant function 27 Continuous 79, 291, 517, 597 Converge 332, 474 Coordinate axes 20 Coordinates 23, 524 Cosh 246 Cosine 124, 128 Critical point 159 Cubic polynomial 191 Curve sketching 171
D
Decreasing 166 Definite integral 305 Degree of angles 112 Degree of polynomial 165 Density 292 Derivative 63, 567 Diderot 525 Differentiable 65, 595
12 INDEX
Dilation 40 Direction 533, 539 Directional derivative 613 Disc 117 Distance 10, 35, 150, 539, 561 , 615 Distance between point and
plane 561 Divergent 474 Domain of a function 16 Dot product 534
E
Ellipse 41 End point 11, 531 Epsilon and delta 502 Equation 30, 36 Equipotential 586 Estimates 431, 435 Euler relation 605 Exist (integral) 332 Exponential function 236, 242, 256,
447 Exponential growth 262 Exponential substitution 379
F
Factorial 103 Fourier coefficients 326 Function 14, 582 Fundamental theorem 308
G
Geometric series 450 Gradient 591, 613 Graph 24, 35, 583 Greater than 6
H
Half closed 11 Hanging cable 394 Higher derivatives 102, 427 Hyperbola 51, 401 Hyperbolic functions 246, 373, 400 Hyperbolic sine and cosine 246, 373
Implicit differentiation 104 Improper integral 329, 383 Increasing 165 Indefinite integral 288, 313
Induction 88 Inequalities 6, 171, 326 Inflection point 189 Inner function 94 Integer 3 Integral 305 Integral test 483 Integration 287 Integration by parts 342 Intermediate value theorm 162, 218,
519 Interval 11 Inverse function 216, 222 Inverse of numbers 5 Isothermal 586
K
Kinetic energy 522
L
Large 181 Least upper bound 502 Left derivative 67 Length of arc 120 Length of curves 120, 391, 402, 578 Level curve 584 Lie on a surface 605 Limits 147,469, 504, 514 Lines 30, 552 Local maximum 162 Local minimum 162 Located vector 531, 532 Logarithm 247, 275, 449 Lower bound 502
M
Mass 306 Maximum 160, 518 Mean value theorem 178 Minimum 161, 518 Moment 419
N
Natural base 241 Negative integer 3 Negative number 6 Newton quotient 63 Newton's law 622 Norm 537 Normal 557 n-tuple 526 Number 4
o Open ball 541 Open interval 11 Open set 586 Order of magnitude 267, 345 Orthogonal 536 Outer function 93
P
Parabola 45, 166, 182 Parallel 34, 532, 558 Parallelogram 528 Parametric curve 397, 552 Parametric line 552 Parametrization 398, 552 Partial derivative 488 Partial fraction 357, 368 Partition 298 Parts 341 Perpendicular 536, 557, 605 Pi 117,459 Piecewise continuous 320 Plane 556 Point in n-space 524 Point of accumulation 514 Polar coordinates 150, 387 Positive integer 3 Positive number 5 Potential energy 622 Potential function 523 Power series 489, 496 Powers 18, 87 Projection 546 Pythagoras theorem 545
Q
Quadratic equation 48 Quadratic formula 48
R
Radians 120 Radius of convergence 491 Rate of change 107 Ratio test 481 Rational function 185, 197 Rational number 4 Real number 4 Related rates 109. 143 Remainder term 430, 435 Riemann sum 299 Right derivative 67 Rolle's theorem 177
INDEX
s Scalar product 534 Schmoo 97 Schwarz inequality 550 Second derivative test 190 Sector 121 Segment 552 Sequence 473 Series 474 Sine 124, 128 Sinh 246 Sketch curves 181 Slope 30, 59 Speed 106, 571 Sphere 541 Square root 8 Squeezing process 72 Straight line 29 Strictly decreasing 166 Strictly increasing 166,217 Substitution 335, 364, 371 Surface 605 Surface of revolution 409
T
Tangent 129 Tangent line 78, 569 Tangent plane 606 Taylor formula 431, 437 Taylor polynomial 428 Torus 385 Triangle inequality 550 Trigonometric functions 148 Trigonometric integrals 347
u Unit vectors 543, 611 Upper bound 502 Upper sum 296, 300
V
Value 14, 483 Vector field 621 Velocity 569 Very large 181 Volume of revolution 381
W
Work 415
13
f cosax 1 . 65. -. - dx = - In Ism axl + C
sm ax a
f " . COS"+I ax 66. cos ax sm ax dx = - (n + I)a + C, n ~-1
f sin ax 1 67. --dx = --lnlcosaxl+C
cos ax a
f . ".. sin"-I ax cos,"+1 ax n - 1 f . "-2 .. 68. sm ax cos ax dx = - + -+ SID ax cos ax dx ,
a(m+n) m n
n ~-m (If n = -m, use No. 86.)
f . " .. sin"+1 ax cos,"-1 ax m - 1 f . " ,"-2 69. SID ax cos ax dx = a(m + n) + m + n SID ax cos ax dx,
m ~ -n (If m = -n, use No. 87.)
n. f d: = -! tan (~ - ~) + C 1 + SID ax a 4 2
iJ. f d: = ! tan (~ + ~) + C 1 - SID ax a 4 2
f dx 2 -1 [~~ ax] 74. = tan -- tan- + C b+ ccosax aVb2 _ c2 b+ c 2 '
75. f dx = 1 In Ic+ bcosax+ v~sinaxl + C b2 < c2 b + c COS ax aV,2 _ b2 b + , COS ax '
f dx 1 ax f dx 1 ax 76. = - tan - + C 77. = - - cot - + C 1 + cos ax a 2 1 - cos ax a 2
78. f x sin ax dx = ~ sin ax - ~ cos ax + C 79. f x cos ax dx = ~ cos ax + ~ sin ax + C
" . x n ,,-1 f "f SO. x SID ax dx = - -; COS ax + a x COS ax dx
" x. n ,,-1. f " f 81. x cos ax dx = -; SID ax - ii x SID ax dx
82. f tan ax dx = - ~ In Icos axl + C
84. f tan 2 ax dx = ~ tan ax - x + C
f 0 tan"-I ax f 0-2 86. tan ax dx = a(n _ 1) - tan ax dx,
f " cot"-I ax f "-2 87. cot ax dx = - a(n _ I) - cot ax dx,
88. f sec ax dx = ~ In Isec ax + tan axl + C
83. f cot ax dx = ~ In Isin axl + C
f 2 1 85. cot ax dx = - a cot ax - x + C
n ~ 1
n ~ 1
89. f esc ax dx = - ~ In Icsc ax + cot axl + C
Continued o~erleaf.
121 90. sec ax dx = ~ tan ax + C I 2 1
91. esc ax dx = - ~ cot ax + C
92 I " dx - sec"-2 ax tan ax + n - 2 I .-2 dx . sec ax - a(n _ 1) n _ 1 sec ax ,
I " esc"-2ax cot ax n- 21 " - 2 93. esc ax dx = - a(n _ 1) + ;; _ 1 esc ax dx,
I . " sec ax 94. sec ax tan ax dx = --+ c,
na n;lliO
I " esc" ax 95. esc ax cot ax dx = - --+ c,
na n;lliO
96. I sin -I ax dx = x sin-I ax + ~ ..; 1 - a2x2 + C
97. leos-I axdx = X COS-I ax - ~";I - a2x2+C
9S. Itan-Iaxdx = xtan-IaX-2~ln(l+a2x2)+C
I " . -I X "+1 . -I a I x·+1 d.r 99. x sm ax dx = -+ 1 sm ax - -- ,
n n + 1 ..; 1 _ a2x2
I " -I x·+1 -I a I X"+I d.r 100. x cos ax dx = -+ 1 cos ax+ -+ '
n n 1 ..; 1 _ a2x2
n ;IIi 1
n;lli 1
n;lli -1
n;lli -1
O "-I d X -I I "+1 1 I. x tan ax x = n+ 1 tan
a I X"+I d.r ax - n + 1 1 + a2x2 ' n ;IIi -1
102. I eoz dx = ~ e"" + c I .. I b·' 103. b dx = ~ In b + C, b> 0, b;lli 1
104. I .. xe·· dx = ~ (ax - 1) + c
a2 105 I
••• dx 1. •• n I .-1 •• dx . xe =-xe -- x e a a
06 I "b·' dx x"b·· n I .-Ibaz dx
1. x =alnb-alnb x , b > 0, b;lli 1
107. f/'sinbxdx = a2~b2(aSinbX-bCOSbX)+C
lOS. I e·' cos bx dx = a2 ~ b2 (a cos bx + b sin bx) + C
109. lin ax dx = x In ax - x + C
I ,,+1 ,,+1 110. 1" Inaxdx = :+ Ilnax - (nx+ 1)2 + c,
lli. I x-lin ax dx = ~ (In ax)2 + C
113. f sinh ax d.r = ~ cosh ax + C
I. 2 sinh 2ax x
liS. smh ax dx = ----:ta - 2 + C
n;lli -1
112. I _Idx = In lin axl + C x nax
114. fCOShaxdx = ~Sinhax+C
I 2 sinh 2ax x
116. cosh ax dx = ----:ta + 2 + C
7 f · h" dx sinh"-I ax cosh ax n - 1 I . h"-2 dx ll. sm ax = --- sm ax, na n
n;lliO
f hn d Coshn-1axSinhax+n- l f h n - 2 d 118. cos ax x = -- cos ax x,
na n n-,l.O
119. fXSinhaxdx = ~coshax-~sinhax+C
120. fx cosh ax dx = ;z: sinh ax - ~ cosh ax + C a a
121. f xn sinh ax dx = ~ cosh ax - ~ f xn - 1 cosh ax dx
122. n x . n "-1. f " f x cosh ax dx = -; smh ax -;;- x smh ax dx
123. f tanh ax dx = ~ In (cosh ax) + C
125. f tanh2 ax dx = x - ~ tanh ax + C
124. f coth ax dx = ~ In Isinh axl + C
126. fcoth 2 ax dx = x - ~cothax+C
f " tanh"-l ax f n-2 127. tanh ax dx = - (n _ l)a + tanh ax dx , n ;Ii 1
f n cothn - 1 ax f "-2 128. coth ax dx = - (n _ 1)a + coth ax dx, n ;Ii 1
129. fsech ax dx = ~ sin -I (tanh ax) + C
130. fCSChax dx = ~In Itanh¥1 +C
132. f CSCh2 ax dx = - ~ coth ax + C
f 2 1 131. sech ax dx = ~ tanh ax + C
33 f h" d sech"-2 ax tanh ax + n - 2 f h,,-2 d I . sec ax x = (n _ 1)a ~ _ 1 sec ax x,
f h" dx csch"-2 ax coth ax n - 2 f h"-2 .J_ 134. csc ax = - - --- csc ax .... , (n - 1)a n - 1
f " sech" ax 135. sech ax tanh ax dx = - --- + C, n;liO
136.
137.
138.
139.
140.
na
f . csch" ax csch ax coth ax dx = - --- + c,
na
fea. sinh bx dx = ea. [~- e-b• J + c,
2 a+b a-b
n;liO
= (n - 1)!, n > O.
a>O
( 12 r / 2 ._, • 2 · 4·6 .. ·n 2
cos x dx =
n -,I. 1
n-,l.l
if n is an even integer ~ 2, 141. Jo sin"xdx=
. { I . 3· 5 · .. (n - 1) 71'
1 2· 4· 6 .. . (n - 1) , 3 · 5·7 .. ·n
if n is an odd integer ~ 3