8.

1 Cauchy’s Integral Formula: Exercises

1. Prove Cauchy’s integral formula for multiply connected regions.
Solution. Let D be a multiply connected domain such that @D = C [ (°C1 ) [ · · · [ (°Cn ),
where C, C1 , · · · , Cn are positively oriented simple closed contours, and C1 , · · · , Cn lie(s)
interior to C. Let z0 2 D be a point. If f (z) is analytic in a domain ≠ containing D then
Z X 1n Z
1 f (z) f (z)
f (z0 ) = dz ° dz.
2ºi C z ° z0 2ºi Ck z ° z0
k=1

Ω
D z0
Γ

-C1
-C2 -C n

Proof. Choose a small number ≤ > 0 such that {z | |z ° z0 | ∑ ≤} Ω D. Let D≤ = D ° {z |

|z ° z0 | ∑ ≤}. For simplicity, we denote by ° the circle {z | |z ° z0 | = ≤} oriented in
f (z)
the positive sense. Since is analytic in ≠ ° {z0 } (which contains the closure D≤ ),
z ° z0
Theorem 7.5 (or Cauchy’s theorem) implies that
Z Z
f (z) f (z)
0= dz = dz
@D≤ z ° z 0 C[(°°)[(°C1 )[···[(°Cn ) z ° z0
Z Z Xn Z
f (z) f (z) f (z)
= dz + dz + dz
z ° z 0 °° z ° z0 z ° z0
C k=1 °Ck
Z Z X n Z
f (z) f (z) f (z)
= dz ° dz ° dz.
C z ° z 0 ° z ° z0 Ck z ° z0
k=1

Then it follows from Cauchy’s integral formula (Theorem 8.1) that

Z Z X 1 n Z
1 f (z) 1 f (z) f (z)
f (z0 ) = dz = dz ° dz.
2ºi ° z ° z0 2ºi C z ° z0 2ºi Ck z ° z0
k=1

////
f (z)
Remark. Since is analytic in ≠ ° {z0 }, we can repeat the above argument to
(z ° z0 )m+1
conclude that
Z Xn Z
m! f (z) m! f (z)
f (z0 ) =
(m)
dz ° dz, m 2 N.
2ºi C (z ° z0 )m+1 2ºi Ck (z ° z0 )m+1
k=1
2. (a) If P (z) is a polynomial of degree n, prove that
Z
P (z)
dz = 0.
|z|=2 (z ° 1)n+2
Z
P (z) 2ºi
Solution. dz = P (n+1) (1) = 0. ( * P (n+1) (z) = 0.)
|z|=2 (z ° 1)n+2 (n + 1)!
(b) If n and m are positive integers, show that
Z Z
(n ° 1)!ez (m ° 1)!ez
dz = dz
C (z ° z0 ) C (z ° z0 )
n m

along any contour C containing z0 .

Solution. Cauchy’s integral formula(Theorem 8.2) implies that, for any n 2 N,
Z
(n ° 1)!ez dn°1 ØØ
dz = 2ºi Ø ez = 2ºiez0 .
C (z ° z0 )
n dz n°1 z=z0

3. Evaluate the following integrals, where C is the circle |z| = 3.

Z Ø
ez Ø
(a) dz = 2ºiez Ø = 2ºie2 .
C z ° 2 z=2
Z 2 Ø
ez 2Ø
(b) dz = 2ºiez Ø = 2ºie4 .
C z ° 2 z=2
Z
d ØØ Ø
2
ez z2 z2 Ø
(c) dz = 2ºi Ø e = 4ºize Ø = 8ºie4 .
C (z ° 2)
2 dz z=2 z=2
Z
ez sin z d ØØ
(d) dz = 2ºi Ø (ez sin z) = 2ºie2 (sin 2 + cos 2).
C (z ° 2)
2 dz z=2
Z
3z 4 + 2z ° 6 2ºi d2 ØØ
(e) dz = 2 Øz=2
(3z 4 + 2z ° 6) = ºi[36z 2 ]z=2 = 144ºi.
C (z ° 2) 3 2! dz
§

4. Use partial fractions to evaluate the following integrals.

(&
⌦o 7.5\
Ò ¶⌃
⌥ ◆6 x◆#
† å⇥ ◆ÿ
‘>⇡€¶√
⇧ ∫e ✓_
î vm⇥
˛ ◆. (d)\
¶s~
⌥ ”Z
ΩO‹ºñ–€¶%
⇧ ✓_
3 vm⇥
˛ ◆.)
Z
1
(a) dz The singularities ±1 are contained in |z| < 2.
|z|=2 z ° 1
2

Z Z Z ≥ 1
1 1 1 1 ¥
dz = dz = ° dz = 0.
|z|=2 z ° 1 |z|=2 (z ° 1)(z + 1) 2 |z|=2 z ° 1 z + 1
2

Z
1
(b) dz The singularities ±1 are contained in |z| < 2.
|z|=2 z2 +1
Z Z ≥ 1
1 1 1 ¥
dz = ° dz = º ° º = 0.
|z|=2 z +1
2 2i |z|=2 z ° i z+i
 Z
1
(c) dz The singularities ±1, ±i are contained in |z| < 2.
|z|=2 z4 ° 1
Z Z ≥ 1
1 1 1 i i ¥
dz = ° ° + dz
|z|=2 z4 ° 1 4 |z|=2 z ° 1 z + 1 z + i z ° i
2ºi
= (1 ° 1 ° i + i) = 0.
4
Z
z 3 + 3z ° 1
(d) dz The singularities 1, °2 are contained in |z| < 3.
|z|=3 (z ° 1)(z + 2)
Z Z ≥ z 3 + 3z ° 1 z 3 + 3z ° 1 ¥
z 3 + 3z ° 1 1
dz = ° dz
|z|=3 (z ° 1)(z + 2) 3 |z|=3 z°1 z+2
2ºi 3
= ([z + 3z ° 1]z=1 ° [z 3 + 3z ° 1]z=°2 ) = 12ºi.
3
§
| (&
⌦o 7.5\
Ò ¶ s6
⌥ xÙ
† «€
 ¶s: 7
⇧ ß8{
· ⌃Ï̄
9 Õ&⇤ì
h ¶s) Fix a small number ≤ > 0 such that
⌅€
⇧

{z | |z ° 1| ∑ ≤} [ {z | |z + 2| ∑ ≤} Ω {z | |z| < 3}

and {z | |z ° 1| ∑ ≤} \ {z | |z + 2| ∑ ≤} = ;. Then Theorem 7.5 implies that

Z Z Z
z 3 + 3z ° 1 z 3 + 3z ° 1 z 3 + 3z ° 1
dz = dz + dz
|z|=3 (z ° 1)(z + 2) |z°1|=≤ (z ° 1)(z + 2) |z+2|=≤ (z ° 1)(z + 2)
≥ z 3 + 3z ° 1 Ø z 3 + 3z ° 1 ØØ ¥
Ø
= 2ºi Ø + Ø = 12ºi.
z+2 z=1 z°1 z=°2

Z
z
5. Evaluate the integral I := dz, where C is the circle
(16 ° C + i) z 2 )(z
z
(Note that the function has three singularities at z = ±4, ° i. Fix a
(16 ° z 2 )(z + i)
constant 0 < ≤ ∑ 1/2. We use Theorem 7.5.)

(a) |z| = 2 : Only one singularity °i is located inside |z| = 2.

Z
z 2ºiz ØØ 2º
Solution. I = dz = 2 Øz=°i
= . ////
|z+i|=≤ (16 ° z 2 )(z + i) 16 ° z 17

(b) |z ° 4| = 2 : Only one singularity 4 is located inside |z ° 4| = 2.

Z Ø
z 2ºiz Ø ºi
Solution. I = dz = ° Ø =° . ////
|z°4|=≤ (16 ° z 2 )(z + i) (4 + z)(z + i) z=4 4 +i

(c) |z + 4| = 2 : Only one singularity °4 is located inside |z + 4| = 2.

Z Ø
z 2ºiz Ø ºi
Solution. I = dz = Ø = . ////
|z+4|=≤ (16 ° z 2 )(z + i) (4 ° z)(z + i) z=°4 4 °i

(d) |z| = 1/2 : No singularities are located inside |z| = 1/2, and Cauchy’s theorem implies
that I = 0. (* z/((16 ° z 2 )(z + i)) is analytic in |z| < 3/4.) ////
 (e) |z| = 5 : Every singularity is contained in the domain |z| < 5.
Solution.
Z Z
z z
I= dz + dz
|z+i|=≤ (16 ° z 2 )(z + i)
|z°4|=≤ (16 ° z 2 )(z + i)
Z
z 2º ºi ºi
+ dz = ° + = 0.
|z+4|=≤ (16 ° z 2 )(z + i) 17 4 + i 4 °i

|¬
“ÏrÏ
⌧ r√
⌧ ∫\
¶ s6
⌥ xÙ
† «€
 ¶s (Optional): Since
⇧

z 4+i 1 4°i 1 i 1
= · ° · ° · ,
(16 ° z 2 )(z + i) 34 z + 4 34 z ° 4 17 z + i

it follows from Cauchy’s integral formula that

Z Z Z Z
z 4+i dz 4°i dz i dz
dz = ° °
|z|=5 (16 ° z )(z + i) 34 |z|=5 z + 4 34 |z|=5 z ° 4 17 |z|=5 z + i
2
≥4 + i 4 ° i i ¥
= 2ºi ° ° = 0.
34 34 17

6. Find the first five coe±cients in the Maclaurin expansion for

(Let O(z n ) denote a quantity z n f (z) where f (z) is bounded near the origin.)

z2 z3 z4
(a) ez sin z : Solution. Since ez = 1 + z + + + + z 5 R1 (z) and sin z =
2 6 24
z3
z° + z 5 R2 (z) for some (continuous) functions R1 (z), R2 (z) near the origin,
6
≥ z2 z3 z4 ¥≥ z3 ¥
ez sin z = 1 + z + + + + z 5 R1 (z) z ° + z 5 R2 (z)
2 6 24 6
3
z
= z + z2 + + O(z 5 ).
3

1 z2 z4
(b) : Solution. Since cos z = 1 ° + ° z 6 R(z) for a (continuous) function
cos z 2 24
R(z) near the origin,
1 1
= ≥ ¥
cos z z2 z4
1 ° 2 ° 24 + z 6 R(z)
≥ z2 z4 ¥ ≥ z2 z4 ¥2
=1+ ° + z 6 R(z) + ° + z 6 R(z) + O(z 6 )
2 24 2 24
1 2 5 4
= 1 + z + z + O(z ). 6
2 24
| ⇥
◆èr €
… ¶s({
⇧ ⌃Ï̄
9 Õ√∫Ü
∆ ì
< ßF⌫ 318A
§: 9Å
· ú 7]
© ⌃)`
X ¶ s6
⌥ xΩ
† … √
+ ∫ e
✓_
î vm⇥
˛ ◆. #
Qã
⌅ ~
" ”Z
Ω O`
¶ Ê
⌥ º8⇤
◆ï∏
power series Ñ
⌅>
⇡ h_ ƒ
ª{9
⌃$⌦\
Ì Å _K
⌫ ⌃ıç
H∞
✏ †_
˙ vm⇥
˛ ◆.
 2
(c) ez+z : Solution. Since z + z 2 = z(1 + z),

2 (z + z 2 )2 (z + z 2 )3 (z + z 2 )4
ez+z = 1 + (z + z 2 ) + + + + O(z 5 )
2 6 24
3 7 25
= 1 + z + z 2 + z 3 + z 4 + O(z 5 ).
2 6 24
z z
(d) e 1°z : Solution. Since = z + z 2 + z 3 + z 4 + z 5 R(z) for some (continuous)
1°z
function R(z), it follows that
2
z
+z 3 +z 4 +z 5 R(z)
e 1°z = ez+z = 1 + (z + z 2 + z 3 + z 4 + z 5 R(z))
+ (z + z 2 + z 3 + z 4 + z 5 R(z))2 + (z + z 2 + z 3 + z 4 + z 5 R(z))3
+ (z + z 2 + z 3 + z 4 + z 5 R(z))4 + O(z 5 )
= 1 + (z + z 2 + z 3 + z 4 + z 5 R(z)) + (z + z 2 + z 3 + z 4 R1 (z))2
+ (z + z 2 + z 3 R2 (z))3 + (z + z 2 R3 (z))4 + O(z 5 )
3 13 73
= 1 + z + z 2 + z 3 + z 4 + O(z 5 ).
2 6 24
§
| Maple`
¶ s6
⌥ x◆#
† å≤ö̇`
¶%
⌥ ⇧`
3 ¶√
⌥ ∫e✓_
î vm⇥
˛ ◆. taylor( exp(z)*sin(z), z=0, 5); \
¶9
⌥ {
ß
⇤◆ì
4 ¶z⇥
¥ù' ⌅ (a)_ ≤
ü◆Ä ö̇`
¶%
⌥ ⇧_
3 ◆. ï
vm⇥
˛ ∏π
ߥ̇
° ò`¶ò
⌥ –9Ä⌅ ?taylor\¶z
⌥ ⇥
¥ùü.
'

7. Suppose f (z) and g(z) are analytic at z0 with

f (z0 ) = f 0 (z0 ) = · · · = f (n°1) (z0 ) = 0, g(z0 ) = g 0 (z0 ) = · · · = g (n°1) (z0 ) = 0.

If g (n) (z0 ) 6= 0, show that

f (z) f (n) (z0 )
lim = (n) .
z!z0 g(z) g (z0 )
This is a generalization of Theorem 5.7.
Solution.1 We use Theorem 8.3. There exist a constant ± > 0 and two functions R1 (z),
R2 (z) which are defined and continuous on N (z0 ; ±) such that
1
X f (k) (z0 )
f (z) = (z ° z0 )k = (z ° z0 )n R1 (z)
k!
k=0
≥ h f (n) (z ) f (n+1) (z ) i¥
0 0
= (z ° z0 )n + (z ° z0 ) + · · ·
n! (n + 1)!

and
1
X g (k) (z0 )
g(z) = (z ° z0 )k = (z ° z0 )n R2 (z)
k!
k=0
≥ h g (n) (z ) g (n+1) (z ) i¥
0 0
= (z ° z0 )n + (z ° z0 ) + · · ·
n! (n + 1)!
1 f (z)¸
< ¶"
⌥
g(z)\ ⇤/
4L√
庖
–Ñ⌅>
⇡ h◆ç
H√
✏ ∫Z
Oìr ◆≈
… “ ◆ö
∏m Ë
H#
⌫å⌧ ˙
<ø∫[
jØπ.
 in N (z0 ; ±). Then it follows that

f (z) (z ° z0 )n R1 (z) R1 (z) R1 (z0 ) f (n) (z0 )

= = ! = (n)
g(z) (z ° z0 ) R2 (z)
n R2 (z) R2 (z0 ) g (z0 )
as z ! z0 . §
|z ¥|
⇥ ⌃&
9 ⇤ì
h ⌅> ⇢Ì
ñ`
fl ¶0
⌥ AK
⌫"
fç
H⇥
✏ ◆6
ßı ∞
£ †s j
˙ t√
˛ ∫e
✓_
î ◆.
vm⇥
˛
f (z) f 0 (z) f (n°1) (z) f (n) (z) f (n) (z0 )
lim = lim 0 = · · · = lim (n°1) = lim (n) = (n) .
z!z0 g(z) z!z0 g (z) z!z0 g (z) z!z0 g (z) g (z0 )

8. Evaluate the following limits, using either the previous exercise or Theorem 8.3.
ez ° 1 ° z ez ° 1 ez 1
(a) lim = lim = lim = .
z!0 z 2 z!0 2z z!0 2 2
X1 n
z z2
§⇥◆èr€
… ¶s: Note that ez =
⇧ = 1+z+ + z 3 R(z) for z 2 C, where R(z) is
n=0
n! 2
continuous at the origin. Therefore the limit is equal to 1/2. ////

sin z cos z
(b) lim = lim = 1. ////
z!0 z ° z3 z!0 1 ° 3z 2
sin z cos z
(c) lim = lim = 1. ////
z!0 ez ° 1 z!0 ez

sin z ° z cos z ° 1 ° sin z

(d) lim = lim = lim = 0.
z!0 cos z ° 1 z!0 ° sin z z!0 ° cos z
X1
(°1)n z 2n+1 z3
§ ⇥
◆èr €
… ¶s: Recall that sin z =
⇧ = z° + z 5 R1 (z) and cos z =
n=0
(2n + 1)! 6
X1
(°1)n z 2n z2 4
= 1 ° z R2 (z), where R1 (z) and R2 (z) are continuous. Therefore the
n=0
(2n)! +
limit is equal to 0. §

9. If f (z) is analytic at z = z0 , and |f (n) (z0 )| ∑ nk for each n (k fixed), show that f (z) is
actually an entire function.
Solution. Recall that n! ∏ (n/e)n for n 2 N.2 Then it follows that
Ø f (n) (z ) Ø1/n nk/n nk/n e(n1/n )k
Ø 0 Ø
Ø Ø ∑ ∑ = !0 as n ! 1.
n! (n!)1/n n/e n
X1
f (n) (z0 )
By Theorem 6.12, the power series converges at every point in the complex
n=0
n!
plane, and hence entire.
X1
f (n) (z0 )
Since f (z) = in a neighborhood of z0 , f (z) is actually entire. (Refer to the
n=0
n!
identity theorem: Theorem 8.15.) §

2 6.4]
⌃
X ⌅_
⇢
É vÎ
˛H]
⌫j€¶s\
⇧ Å≈“$
⇤‹
3ºñ–e
✓6
î ß.
£
11. Suppose f (z) and g(z) are analytic in a simply connected domain D. Prove that
Z z1 Z z1
f (z)g 0 (z)dz = f (z1 )g(z1 ) ° f (z0 )g(z0 ) ° f 0 (z)g(z)dz,
z0 z0

where the path of integration is any contour from z0 to z1 that lies in D.

Solution. Since f 0 (z)g(z) and f (z)g 0 (z) are both analytic in D, the integrals are independent
of the choice of a contour from z0 to z1 . Integrate the both sides of the following identity,
and then apply Theorem 7.7.
d
(f (z)g(z)) = f 0 (z)g(z) + f (z)g 0 (z).
dz
§

12. Suppose f (z) is continuous, but not necessarily analytic, on a contour C. Show that the
function Z
f (≥)
F (z) = d≥
C ≥ °z
is analytic at each z not on C, with
Z
f (≥)
F 0 (z) = d≥.
C (≥ ° z)2

Solution. Follow the proof of Morera’s theorem. §

13. Choose a specific determination, find Maclaurin expansions for the following functions, and
state the region for which the expansion is valid.

(a) (1 + z)Æ , 0 < Æ < 1 (Principal branch)

Solution. (1 + z)Æ (Principal branch) is defined by (1 + z)Æ = eÆLog(1+z) , z 6= 0.
Since Log(1 + z) is analytic in D = C ° {x 2 R | x ∑ °1}, eÆLog(1+z) is also analytic
in D. Then Theorem 8.3 shows that (1 + z)Æ can be expressed as a Taylor series in
|z| < 1.3 Note that

d Æ
f 0 (z) = eÆLog(1+z) [ÆLog(1 + z)] = eÆLog(1+z) = Æ(1 + z)Æ°1 , |z| < 1
dz 1+z
..
.
√ !
Æ
f (n)
(z) = Æ(Æ ° 1) · · · (Æ ° n + 1)(1 + z) Æ°n
= n! (1 + z)Æ°n , |z| < 1,
n
√ !
Æ Æ(Æ ° 1) · · · (Æ ° n + 1)
where = .
n n!
1
√ !
X Æ n
Therefore (1 + z) =
Æ
z for |z| < 1. (binomial series) §
n=0
n

3 Note that N (0; 1) = {z | |z| < 1} is the largest open disk contained in D.