CHAPTER 10

Exercise 10.1
1. (a) Yes.
(b) Yes, because at t = 0, the value of y for the two functions are identical: 3
0
= 1, and
3
2(0)
= 1.
(a) Yes.
(b) No, because at t = 0, the value of y for the two functions are unequal: 4
0
= 1, but
3

4
0

= 3.
2. (a) Let w = 5t (so that dw/dt = 5), then y = e
w
and dy/dw = e
w
. Thus,by the chain rule,
dy
dt
=
dy
dw
dw
dt
= 5e
w
= 5e
5t
.
(b) Let w = 3t , then y = 4e
w
and dy/dw = 4e
w
. Thus,we have
dy
dt
=
dy
dw
dw
dt
= 12e
3t
(c)Similarly to (b) above, dy/dt = 12e
2t
.
3. The rst two derivatives are y
0
(t) = y
00
(t) = e
t
= (2.718)
t
. The value of t can be either
positive, zero, or negative. If t > 0, then e
t
is clearly positive; if t = 0, then e
t
= 1, again
positive; nally, if t < 0, say t = 2, then e
t
= 1/ (2.718)
2
, still positive. Thus y
0
(t) and y
00
(t)
are always positive, and the function y = e
t
always increases at an increasing rate.
4. (a) The curve with a = 1 is the mirror image of the curve with a = 1 with reference to the
horizontal exis.
(b) The curve with c = 1 is the mirror image of the curve with c = 1 with reference to the
vertical exis.
Exercise 10.2
1. (a) e
2
= 1+2+
1
2
(2)
2
+
1
6
(2)
3
+
1
24
(2)
4
+
1
120
(2)
5
+
1
720
(2)
6
+
1
5040
(2)
7
+
1
40320
(2)
8
+
1
362880
(2)
9
+
1
3628800
(2)
10
= 1 + 2 + 2 + 1.333 + 0.667 + 0.267 + 0.089 + 0.025 + 0.006 + 0.001 + 0.000 = 7.388
(b) e
1/2
= 1+
1
2
+
1
2
(
1
2
)
2
+
1
6
(
1
2
)
3
+
1
24
(
1
2
)
4
+
1
120
(
1
2
)
5
= 1+0.5+0.125+0.021+0.003+0.000 = 1.649
56
2. (a) The derivatives are:
0
= 2e
2x
,
00
= 2
2
e
2x
,
0
0
0
0
= 2
3
e
2x
, or in general
(k)
= 2
k
e
2x
. Thus
we have
0
(0) = 2,
00
(0) = 2
2
, or more generally
(k)
(0) = 2
k
. Accordingly,
P
n
= 1 + 2x +
1
2!
2
2
x
2
+
1
3!
2
3
x
3
+ +
1
n!
2
n
x
n
= 1 + 2x +
1
2!
(2x)
2
+
1
3!
(2x)
3
+ +
1
n!
(2x)
n
(b) R
n
=

(n+1)
(p)
(n+1)!
x
n+1
=
2
(n+1)
e
2p
(n+1)!
x
n+1
=
e
2p
(n+1)!
(2x)
n+1
It can be veried that R
n
0 as n .
(c) Hence (x) can be expressed as an innite series:
(x) = 1 + 2x +
1
2!
(2x)
2
+
1
3!
(2x)
3
+
3. (a) $70e
0.04(3)
= $70e
0.12
(b) $690e
0.05(2)
= $690e
0.10
4. (a) 0.07 ( or 7%) (b) 0.03 (c) 0.40 (d) 1 ( or 100% )
5. When t = 0, the two functions have the same value ( the same y intercept ). Also, y
1
= Ae
r
when t = 1, but y
2
= Ae
r
when t = 1. Generally, y
1
= y
2
whenever the value of t in one
function is the negative of the t value in the other; hence the mirror- image relationship.
Exercise 10.3
1. (a) 4 (b) -4 (c) 4 (d) 5
2. (a) 7 (b) -4 (c) -3 (d) -2 (e) 6 (f) 0
3. (a) log
10
(100)
13
= 13 log
10
100 = 13(2) = 26
(b) log
10
(
1
100
) = log
10
1 log
10
100 = 0 2 = 2
(c) ln
3
B
= ln3 lnB
(d) ln Ae
2
= lnA = lne
2
= lnA + 2
(e) lnABe
4
= lnA + lnB + lne
4
= lnA + lnB 4
(f) (log
4
e)(log
e
64) = log
4
64 = 3
4. (a) and (c) are valid; (b) and (d) are not.
5. By denition, e
ln(u/v)
=
u
v
. But we can also write
u
v
=
e
ln u
e
ln v
= e
(ln uln v)
. Equating the two
expressions for
u
v
, we obtain ln
u
v
= lnu lnv.
57
Exercise 10.4
1. If r = 0, then y = Ae
rt
= Ae
0
= A, and the function degenerates into a constant function.
The nonzero requirement serves to preclude this contingency.
2. The graphs are of the same general shap as in Fig. 10.3; the y intercepts will be A (i.e., y = A)
for both.
3. Since y = ab
ct
, we have log
b
y = log
b
a + ct log
b
b = log
b
a + ct.
Thus, by solving for t, we get
t =
log
b
ylog
b
a
c
(c 6= 0)
This is the desired inverse function because it expresses t in terms of y.
4. (a) a = 1, b = 8, and c = 3; thus r = 3 ln8, and y = e
(3 ln 8)t
. We can also write this as
y = e
6.2385t
.
(b) a = 2, b = 7, and c = 2; thus r = 2 ln7, and y = 2e
(2 ln 7)t
. We can also write this as
y = 2e
3.8918t
.
(c) a = 5, b = 5, and c = 1; thus r = ln5, and y = 5e
(ln 5)t
. We can also write this as
y = 5e
1.6095t
.
(d) a = 2, b = 15, and c = 4; thus r = 4 ln15, and y = 2e
(4 ln 15)t
. We can also write this as
y = 2e
10.8324t
.
5. (a) a = 1, b = 7, c = 1; thus t =
1
ln 7
lny(=
1
1.9459
lny = 0.5139 lny)
(b) a = 1, b = 8, c = 3; thus t =
1
ln 8
ln3y(=
1
2.0795
ln3y = 0.4809 ln3y)
(c) a = 3, b = 15, c = 9; thus t =
3
ln 15
ln9y(=
3
2.7081
ln9y = 1.1078 ln9y)
(d) a = 2, b = 10, c = 1; thus t =
2
ln 10
lny(=
2
2.3026
lny = 0.8686 lny)
6. The conversion involved is Ae
rt
= A(1+
i
c
)
ct
, where c represents the number of compoundings
per year. Similarly to formula (10.18), we can obtain a general conversion formula r =
c ln(1 +
i
c
).
(a) c = 1, and i = 0.05; thus r = ln1.05.
(b) c = 2, and i = 0.05; thus r = 2 ln1.025.
(c) c = 2, and i = 0.06; thus r = 2 ln1.03.
58
(d) c = 4, and i = 0.06; thus r = 4 ln1.015.
7. (a)The 45
0
line drawn through the origin serves as a mirror.
(b)Yes, the horizonal axis is a mirror.
(c) Yes, the horizonal axis is a mirror.
Exercise 10.5
1. (a) 2e
2t+4
(b) 9e
17t
(c) 2te
t
2
+1
(d) 10te
2t
2
(e) (2ax +
b)e
ax
2
+bx+c
(f)
dy
dx
= x
d
dx
e
x
+ e
x dx
dx
= xe
x
+ e
x
= (x + 1)e
x
(g)
dy
dx
= x
2
(2e
2x
) + 2xe
2x
= 2x(x + 1)e
2x
(h)
dy
dx
= a(xbe
bx+c
+ e
bx+c
) = a(bx + 1)e
bx+c
2. (a)
d
dt
lnat =
d
dt
(lna + lnt) = 0 +
d
dt
lnt =
1
t
.
(b)
d
dt
lnt
c
=
d
dt
c lnt = c
d
dt
lnt =
c
t
.
3. (a)
dy
dt
=
35t
4
7t
5
=
5
t
(b)
dy
dt
=
act
c1
at
c
=
c
t
(c)
dy
dt
=
1
t + 19
(d)
dy
dt
= 5
2(t+1)
(t+1)
2
=
10
t+1
(e)
dy
dx
=
1
x

1
1+x
=
1
x(1+x)
(f)
dy
dx
=
d
dx
[lnx + 8 ln(1 x)] =
1
x
+
8
1x
=
19x
x(1x)
(g)
dy
dx
=
d
dx
[ln2x ln(1 + x)] =
2
2x

1
1+x
=
1
x(1+x)
(h)
dy
dx
= 5x
4 2x
x
2
+ 20x
3
lnx
2
= 10x
3
(1 + 2 lnx
2
) = 10x
3
(1 + 4 lnx)
4. (a)
dy
dt
= 5
t
ln5 (b)
dy
dt
=
1
(t+1) ln 2
(c)
dy
dt
= 2(13)
2t+3
ln13
(d)
dy
dx
=
14x
7x
2
1
ln 7
=
2
xln 7
(e)
dy
dx
=
16x
(8x
2
+3) ln 2
(f)
dy
dx
= x
2 d
dx
log
3
x + log
3
x
d
dx
x
2
= x
2 1
xln 3
+ (log
3
x)(2x) =
x
ln 3
+ 2xlog
3
x
5. (a) Let u = f(t), so that du/dt = f
0
(t). Then
d
dt
b
f(t)
=
db
u
dt
=
db
u
du
du
dt
= (b
u
lnb)f
0
(t) = f
0
(t)b
f(t)
lnb
(b) Let u = f(t). Then
d
dt
log
b
f(t) =
d
dt
log
b
u =
d
du
log
b
u
du
dt
=
1
uln b
f
0
(t) =
f
0
(t)
f(t)
1
ln b
59
6. For V = Ae
rt
, the rst two derivatives are V
0
= rAe
rt
> 0 and V
00
= r
2
Ae
rt
> 0
Thus V is strictly increasing at an increasing rate, yielding a strictly convex curve. For A =
V e
rt
, the rst two derivatives are
A
0
= rV e
rt
< 0 and A
00
= r
2
V e
rt
> 0
Thus A is strictly decreasing at an increasing rate (with the negative slope taking smaller
numerical values as t increases ), also yielding a strictly convex curve.
7. (a) Since lny = ln3x ln(x + 2) ln(x + 4), we have
1
y
dy
dx
=
1
x

1
x+2

1
x+4
=
8x
2
x(x+2)(x+4)
and
dy
dx
=
8x
2
x(x+2)(x+4)
3x
(x+2)(x+4)
=
3(8x
2)
(x+2)
2
(x+4)
2
(b) Since lny = ln(x
2
+ 3) + x
2
+ 1, we have
1
y
dy
dx
=
2x
x
2
+3
+ 2x =
2x(x
2
+4)
x
2
+3
and
dy
dx
=
2x(x
2
+4)
x
2
+3
(x
2
+ 3)e
x
2
+1
= 2x(x
2
+ 4)e
x
2
+1
Exercise 10.6
1. Since A = Ke
2

trt
, we have lnA = lnK + 2

t rt. Dierentiation with respect to t yields

1
A
dA
dt
= t
1/2
r or
dA
dt
= A(t
1/2
r)
Setting
dA
dt
= 0, we then nd: t

= 1/r
2
.
In the second derivative,
d
2
A
dt
2
= A
d
dt
(t
1/2
r) + (t
1/2
r)
dA
dt
the second term vanishes when
dA
dt
= 0. Thus
d
2
A
dt
2
= A/2

t
3
< 0, which satises the
second-order condition for a maximum.
2.
d
2
A
dt
2
= A
d
dt
(
ln 2
2

t
r) + (
ln 2
2

t
r)
dA
dt
= A
d
dt
(
ln 2
2
t
1/2
r) + 0 =
Aln 2
4

t
3
< 0 [ since A < 0 and
ln2 > 0]
Thus the second-order condition is satised.
3. (a) Since A = V e
rt
= f(t) e
rt
, we have lnA = lnf(t) rt, and
1
A
dA
dt
=
f
0
(t)
f(t)
r = r
v
r
or
dA
dt
= A(r
v
r)
Inasmuch as A is nonzero, dA/dt = 0 if and only if r
v
= r.
(b) The second derivative is
d
2
A
dt
2
= A
d
dt
f
0
(t)
f(t)
= A
d
dt
r
v
< 0 i
d
dt
r
v
< 0
4. The value of t

depends only on the parameter r. Since

dt

dr
=
d
dr
1
4
r
2
=
1
2r
3
< 0, a higher
interest rate means a smaller t

(an earlier optimal time of sale).

60
Exercise 10.7
1. (a) lny = ln5 + 2 lnt; thus r
y
=
d
dt
lny =
2
t
.
(b) lny = lna + c lnt; thus r
y
= c/t.
(c) lny = lna + t lnb; thus r
y
= lnb.
(d) Let u = 2
t
and v = t
2
. Then r
u
= ln2, and r
v
= 2/t. Thus r
y
= r
u
+ r
v
= ln2 + 2/t.
Alternatively, we can write lny = t ln2 + 2 lnt; thus
r
y
=
d
dt
lny = ln2 =
2
t
(e) Let u = t and v = 3
t
. Then r
u
=
d(ln u)
dt
=
d(ln t)
dt
=
1
t
, and r
v
=
d(ln v)
dt
=
d(ln 3
t
)
dt
=
d(t ln c)
dt
= ln3. Consequently, r
y
= r
u
r
v
=
1
t
ln3.
Alternatively, we can write lny = lnt t ln3; thus r
y
=
d
dt
lny =
1
t
ln3
2. lnH = lnH
0
+ bt ln2; thus r
H
= b ln2. Similarly, lnC = lnC
0
+ at lne; thus r
C
= a lne = a.
It follows that r
(C/H)
= r
C
r
H
= a b ln2.
3. Taking log, we get lny = k lnx. Dierentiating with respect to t, we then obtain r
y
= kr
x
.
4. y =
u
v
implies lny = lnu lnv; it follows that r
y
=
d
dt
lny =
d
dt
lnu
d
dt
lnv = r
u
r
v
.
5. By denition, y = Y/P. Taking the natural log, we have lny = lnY lnP. Dierentiation of
lny with respect to time t yields
d
dt
lny =
d
dt
lnY
d
dt
lnP, which means r
y
= r
Y
r
P
where
r
P
is the rate of ination.
6. z = u v implies lnz = ln(u v); thus r
z
=
d
dt
lnz =
d
dt
ln(u v) =
1
uv
d
dt
(u v) =
1
uv
d
dt
[f
0
(t) g
0
(t)] =
1
uv
(ur
u
vr
v
)
7. lnQ
d
= lnk nlnP. Thus, by (10.28),
d
= n, and |
d
| = n.
8. (a) Since lny = lnw + lnz, we have
yx
=
d(ln y)
d(ln x)
=
d(ln w)
d(ln x)
+
d(ln z)
d(ln x)
=
wx
+
zx
(b) Since lny = lnu lnv, we have
yx
=
d(ln y)
d(ln x)
=
d(ln u)
d(ln x)

d(ln v)
d(ln x)
=
ux
+
vx
9. Let u = log
b
y, and v = log
b
x (implying that x = b
v
). Then
du
dv
=
du
dy
dy
dx
dx
dv
=
1
y
(log
b
e)
dy
dx
b
v
lnb.
Since log
b
e =
1
ln b
, and since b
v
= x, we have
du
dv
=
x
y
dy
dx
=
yx
.
10. Since M
d
= f[Y (t), i(t)], we can write the total derivative
dM
d
dt
= f
y
dy
dt
+ f
i
di
dt
.
61
Thus the rate of growth of M
d
is r
M
d
=
dM
d
/dt
M
d
=
f
Y
f
dY
dt
+
f
i
f
di
dt
=
f
Y
f
Y
Y
dY
dt
+
f
i
f
i
i
di
dt
=
f
Y
Y
f
(
1
Y
dY
dt
) +
fii
f
(
1
i
di
dt
) =
M
d
Y
r
Y
+
M
d
i
r
i
.
Alternatively, using logarithms, we may write r
M
d
=
d
dt
lnM
d
=
1
M
d
d
dt
M
d
, but this then leads
us back to the preceding process.
11. By the same procedure used in 10 above, we can nd that r
Q
=
QK
r
K
+
QL
r
L
.
62