Chapter 7

Exercise VII.4.1
Derive the equality
∞
1 − b + x2
Z
dx = π
−∞ (1 − b + x2 )2 + 4bx2

where 0 < b < 1. What goes wrong if b > 1? √

Let γa be the curve around the rectangle with vertices ±a, ±a + i b
(a > 0). Z
1
We can then write 2
dz as follows.
γa 1 + z

Z a Z a Z √b  
1 1 i i
dt − √ dt + − dt
1 + t2 −a 1 + (t + i b)
2 1 + (a + it)2 1 + (−a + it)2
| −a {z } | {z } |
0
{z }
I1 I2 I3

a
Notice I1 = tan−1 (t) → π as a → ∞ and
−a

a a a
√
1 − b + t2 − i2 bt
Z Z Z
dt dt
√ = √ = dt.
−a 1 + (t + i b)2 −a
2
1 + t + i2 bt − b −a (1 − b + t2 )2 + 4bt2

1
Since b < 1, is holomorphic on the (convex) rectangle enclosed by γa .
Z 1 + z2
1
So, 2
dz = 0 and if as a → ∞ I3 → 0 and we set the real part of I1
γa 1 + z
equal to the real part of I2 , then
Z ∞
1 − b + t2
π= 2 2 2
.
−∞ (1 − b + t ) + 4bt

To show that I3 → 0 as a → ∞ we simplify and approximate the integral as

follows.

1
 √
Z b  
1 1
I3 = i − dt
0 1 + (a + it)2 1 + (−a + it)2
√
Z b  
1 1
=i − dt
0 1 + a2 − t2 + i2at 1 + a2 − t2 − i2at
√
b
a2(− t2 − i2at) − (( a2(− t2 + i2at)
Z (( ((
((
1+(( 1+((
=i dt
0 (1 + a2 − t2 )2 + (2at)2
√
Z b
4at
= dt
0 (1 + − t2 )2 + (2at)2
a2
√ 4at
|I3 | ≤ b max √ 2 2 2 2
t∈[0, b] (1 + a − t ) + (2at)

The denominator is a degree 4 polynomial in a, so as a → ∞, |I3 | → 0.

2
Exercise VII.4.2
Let a be a positive number. Evaluate the integrals
Z ∞ Z ∞
1 t2
dt, .
0 t4 + a4 0 t4 + a4
π
Let γr be the boundary of { z : 0 ≤ |z| ≤ r, 0 ≤ Arg z ≤ 4
}.

r Z π/4 Z 0 iπ/4
ireit
Z Z
1 1 e
dz = dt + dt + dt
γr z 2 + a2 t2 + a2 r2 ei2t + a2 2
r it + a
2
|0 {z } | 0
{z } | {z }
I1 I2 I3

Since γr encloses a convex region where 1/(z 2 + a2 ) is holomorphic, the

left hand side of the above equation is 0.
r
I1 = a1 tan−1 ( at ) = a1 tan−1 ( ar ) → 2a
π
as r → ∞.
0

π/4
ireit
Z
|I2 | ≤ dt
0 r2 ei2t + a2
π/4
|r|
Z
≤ dt
0 ||r2 | − |a2 ||
π |r|
= → 0 as r → ∞
4 ||r | − |a2 ||
2

Z r
iπ/4 1
I3 = −e dt
0 + a2 it2
Z r
iπ/4 −it2 + a2
= −e dt
0 t4 + a4
| {z }
I∗
Z r Z r
t2 2 1
I∗ = −i 4 4
dt + a 4 4
dt
0 t +a 0 t +a
So, we have as r → ∞ the following.
Z ∞ Z ∞
π −iπ/4 t2 2 1
e = −i 4 4
dt + a dt
2a 0 t +a 0 t + a4
4
√ √ !
π 2 2
= −i
2a 2 2

3
Separating the real and imaginary parts we find,
√ Z ∞ √ Z ∞
π 2 1 π 2 t2
= dt and = dt.
4a3 0 t4 + a4 4a 0 t4 + a4

4
Exercise VII.5.1
Let f and C beZ as in the statement of Cauchy’s formula for a circle. What
f (ζ)
is the value of dζ when z is in the exterior of C.
C ζ −z
f is holomorphic on C and since z is outside of this disc C encloses,
f (ζ)
is also holomorphic there. So, by Cauchy’s theorem for a convex region
Zζ − z
f (ζ)
dζ = 0.
C ζ −z

5
VII.6.1
Let f be holomorphic in the disk |z − z0 | < R and Cr be the counterclockwise
circle with center z0 and radius r. Prove
ZZ
1
f (z0 ) = 2 f (x + iy)dxdy, 0 < r < R.
πr |z−z0 |<r

Fubini’s theorem can be applied since f is continuous and bounded on

the disc { z ∈ C : |z| ≤ r}.
Z r Z 2π Z r
1 it 1 2 1 2
f (z0 + ρe )ρdθdρ = 2πf (z0 )ρdρ = f (z0 ) r = f (z0 )
πr2 0 0 πr2 0 r2 2

6
Exercise VII.7.1
Find explicity the Cauchy integral of the constant function 1 over the interval
[0, 1]. Z 1
1
The Cauchy integral is f (z) = dt, where z ̸∈ [0, 1].
0 t−z
Let z = x + iy.
Z 1
1
f (z) = dt
0 t − (x + iy)
Z 1
1 t − x + iy
= · dt
0 t − (x + iy) t − x + iy
Z 1
t−x y
= 2 2
+i dt
0 (t − x) + y (t − x)2 + y 2
Z 1−x
u y
= 2 2
+i 2 du
−x u +y u + y2
1 u=1−x
= ln u2 + y 2 + i tan−1 (u/y)
2 u=−x

7
Exercise VII.8.1
Evaluate the following integrals, where C is the counterclockwise unit circle.
Z
sin(z)
1. 38
dz
C z

2πi d37
Z
sin(z) 2πi 2πi
37+1
dz = 37
sin(z) = cos(0) =
C (z − 0) 37! dz z=0 37! 37!
Z  3
z−2
2. dz
C 2z − 1
3 3
(z − 2)3
Z  
z−2 1 z−2
Z Z
1
dz = dz = dz
C 2z − 1 C 8 z − 21 8 C (z − 12 )2+1
1 2πi d2
  
3 πi 1 9πi
= 2
(z − 2) = 6 −2 =−
8 2! dz z=1/2 8 2 8

8
Exercise VII.8.2
Z 2π
1 1
For 0 < r < 1, evaluate dθ.
2π 0 1 − 2r cos(θ) + r2
Let z = eiθ and γ be the counterclockwise unit circle. The integral then
can be written as follows.
Z 2π Z 2π
1 1 1 1
dt = dt
2π 0 1 − 2r cos(t) + r 2 2π 0 1 − re − re−it + r2
it
Z
1 1
= dz
2πi γ z (1 − rz − r/z + r2 )
Z
1 1
= dz
2πi γ (z − rz − r + zr2 )
2

−1 1
Z
1
= · 2
dz
r 2πi γ (z − z/r − rz + 1)
−1 1
Z
1
= · dz
r 2πi γ (z − r)(z − 1r )
−1 1 1/(z − 1r )
Z
= · dz
r 2πi γ z − r
| {z }
Cauchy integral
−1 1
= · 1
r r− r

9
Exercise VII.8.3
Let a, b ∈ C where |a| < 1 < |b| and C be the counterclockwise unit circle.
For all m, n ∈ Z, evaluate

(z − b)m
Z
1
dz.
2πi C (z − a)n

Let f (z) = (z − b)m . Note that for any m ∈ Z and |z| ≤ 1, (z − b) ̸= 0.

So, f is holomorphic on the unit disc whether m < 0 or m > 0 meaning, for
n > 0 we can express the integral as

(a − b)m
Z
f (a) 1 f (z)
= = dz.
(n − 1)! (n − 1)! 2πi C (z − a)n

Now, if n ≤ 0 and m ≥ 0, the integral is over a degree m − n polynomial

which by Cauchy’s theorem for a convex region implies the integral is 0.
Finally, if n < 0 and m < 0, then the integrand is the quotient of two
polynomial where the denominator is never zero on or in the unit disc mean-
ing, the integrand is holomorphic here. So, again by Cauchy theorem for a
convex region, the integral is 0.

10
Exercise VII.9.1
If f (z) = ∞ n
P
n=0 an z , what is the taylor series about the origin of the function
(1 − z)−1 f (z)?
Notice that the taylor series expansion about the origin of (1 − z)−1 is
X∞
z n . So, using the Cauchy-Product the taylor series about the origin of
n=0
∞ X
X n
(1 − z)−1 f (z) is ak z n .
n=0 k=0

11
Exercise VII.10.1
R
Let G ⊆ C be open and f : G → C be continuous. Prove that if R f (z)dz = 0
for any rectangle R ⊆ G whose edges are parallel to the coordinate axes, then
f is holomorphic.
Suppose z0 ∈ G, since G is open there is an open disc D ⊆ G centered at
z0 . Define a function g(z) in a subset of D that is the integral of f along a
2 part curve traveling from z0 to z first horizontally then vertically.
Let z1 be another point and γ∗ be another 2 part curve that connects z
to z1 such that it forms a rectangle R in conjunction with curves connecting
z0 to z and z0 to z1 .

γ∗
z
z1

z0

We can then write

Z 
Z
f  =
(ζ)dζ g(z) + f (ζ)dζ − g(z1 ).
R γ∗
g(z1 ) − g(z)
Z
1
− f (z) = f (ζ)dζ − f (z)
z1 − z z1 − z γ∗
Z
1
= (f (ζ) − f (z)) dζ
z1 − z γ∗
g(z1 ) − g(z)
Z
1
− f (z) ≤ (f (ζ) − f (z)) dζ
z1 − z z1 − z γ∗
1
≤ max |f (ζ) − f (z)| L(γ∗ )
|z1 − z| ζ∈γ∗
1
≤ max |f (ζ) − f (z)| |z1 − z|
|z1 − z| ζ∈γ∗
= max |f (ζ) − f (z)|
ζ∈γ∗

12
 Since f is continous and γ∗ is smooth for any ϵ > 0 there exists a δ > 0
such that maxζ∈γ∗ |f (ζ) − f (z)| < ϵ, which means that f is the complex
derivative of g at z and since z was arbitrary, f is holomorphic in G.

13
Exercise VII.10.2
Let the function f be continuous on { z : |z| < 1, Im z ≥ 0 }, real valued
on the segement −1, 1, and holomorphic on D+ = { z : |z| < 1, Im z > 0 }.
Prove f can be extended holomorphically to the open unit disk.
Let D+ = { z : |z| < 1, Im z > 0 } and D− = { z : |z| < 1, Im z < 0 }. As
seen in exercise 2.8.2, ϕ : D− 7→ C is holomorphic where ϕ(z) = f (z).
We then define the extended version of f as follows.
(
f (z) if z ∈ D+ ∪ [0, 1]
g(z) =
f (z) if z ∈ D−

Let R be a rectangle in the unit disc that has sides parallel to the coordinate
axis and overlaps the real axis. We then consider the ”subrectangle” Rh
that overlaps the real axis in the same way as R, except has height h and is
symmetric about the real axis.

γ1
Rh
h
γ2
w

Z Z
We claim that g(z)dz = h(z)dz, because we can write
R Rh
Z Z Z Z
g(z)dz = g(z)dz + g(z)dz + g(z)dz.
R γ1 Rh γ3

γ1 and γ2 are countours on the two other subrectangles that appear when
drawing Rh , which are contained entirely in (convex) parts of the unit disc
where we already know h is holomorphic.
The edge case where the bottom or top of R intersects the real axis need
to be handled, but first lets work through when R overlaps as in the
Z picture.
Let x0 + iy0 denote the bottom left of R. We can then write h(z)dz
Rh

14
as follows.
Z h/2 Z w
f (x0 − it)(−i)dt + f (x0 − ih/2 + t)dt
0 0
Z h/2 Z h/2
+ f (x0 + w + i(t − h/2))(i)dt + f (x0 + w + it)(i)dt
0 0
Z w Z h/2
+ f (x0 + ih/2 + (w − t))(−1)dt + f (x0 + i(h/2 − t))(−i)dt
0 0
Taking the modulus, applying
Z the triangle inequality, then approximating
each integral we find that h(z)dz is less than or equal to the following.
Rh

h
max f (x0 − it)(−i) + · · · + f (x0 + i(h/2 − t))(−i)
2 t∈[0,h/2]
+ w max f (x0 − ih/2 + t) + f (x0 + ih/2 + (w − t))(−1)
t∈[0,w]

Since h is continuous, if we make h arbitrarily small we find the first term

goes to 0 which leaves us with
w max f (x0 + t) − f (x0 + (w − t)) = w max f (x0 + t) − f (x0 + t) .
t∈[0,w] t∈[0,w]
Z
Since f is real valued we are left with 0 meaning, h(z)dz goes to 0 as h
Rh
goes to 0.
Now to handle the case where the top or bottom of R coincides with the
real axis. Without a loss of generality suppose the bottom of R coincides
with the real axis.
If we form another rectangle Rh that has the same bottom as R except has
height h, then essentially the same argument as before implies the integral
of g over R is the same as on Rh .

h Rh
w

15
Let x0Zbe the lower left of Rh , w it’s width, and h it’s height. We can then
write g(z)dz as follows.
Rh

Z w Z h
g(x0 + t)dt + g(x0 + w + it)(i)dt
0 0
Z w Z h
+ g(x0 + ih + (w − t))(−1)dt + g(x0 + i(h − t))(−i)dt
0 0
Z w
= (g(x0 + t) − g(x0 + ih + (w − t))) dt
0
Z h
+i (g(x0 + w + it) − g(x0 + i(h − t))) dt
0
Both of these integrals go to 0 as h goes to 0 since the first integral just
integrates over the interval [0, w] forward and backward while the second
integral can be approximated by

h max |g(x0 + w + it) − g(x0 + i(h − t))| .

t∈[0,h]

Since g is continuous, as h goes to 0 the max approaches a constant. So the

limit becomes 0 · constant = 0.
Hence the integral of g on any rectangle is 0 which by the previous exercise
implies g is holomorphic.

16
Exercise VII.11.1
Say f is entire and for some n ∈ Z+ , |f (z)/z n | is bounded for |z| > R. Prove
f is a polynomial of degree at most n.
For any z ∈ C, there exists an r > max{|z| , 2R}. Let Cr denote the circle
centered at 0 with radius r.
Since f is holomorphic, we can write

(n − 1)!
Z
(n−1) f (ζ)
f (z) = n
dζ.
2πi Cr (z − ζ)

Since the integral is bounded when |z − ζ| > R, the (n − 1)th derivative

of f is constant, which implies that f is at most a degree n polynomial.

17
Exercise VII.11.2
Let f be a holomorphic map of the open unit disk into itself. Prove that
1
|f ′ | (z) ≤
1 − |z|

for all z in the disk.

Let Cϵ denote the circle with radius ϵ and center z. Since f is holomorphic
we can write
Z
′ 1 f (ζ)
|f (z)| = dζ
2π Cϵ (ζ − z)2
Z
1 f (ζ)
≤ dζ
2π Cϵ (ζ − z)2
|f (ζ)|
Z
1
= dζ
2π Cϵ ϵ2
1 1
≤ 2πϵ
2π ϵ2
1
= .
ϵ
ϵ can be taken arbitrarily close to 1 − |z| meaning,
1
|f ′ (z)| ≤ .
1 − |z|

18
Exercise VII.11.3
Prove that an entire function with a positive real part is constant.
Suppose f is a non-constant entire function with a strictly postive real
part. Since f is holomorphic, for any closed contour γ ⊂ C, we must have
Z
f (z)dz = 0.
γ

However, Z Z
0 < Re f (z)dz ≤ f (z)dz .
γ γ

19
Exercise VII.13.1
Prove that there is no holomorphic function f in the open unit disk such
that f ( n1 ) = 2−n for n = 2, 3, . . ..
Suppose such an f exists. Since f is continuous f (lim( n1 )) = lim(2−n ) ⇒
f (0) = 0. f is non-constant so 0 cannot be infinite order. Suppose 0 is order
m, then we can write for some holomorphic function g.

f (z) = z m g(z)

But then we have

1 1 1 nm 1
= g( ) ⇔ = g( ).
2n nm n 2n n
Again g is holomorphic meaning g(0) = lim(nm /2n ). Contradicting that
g(0) ̸= 0.

20
Exercise VII.13.2
Let f be a holomorphic function in the open subset G of C. Let the point z0
1
of G be a zero of f of order m. Prove that there is a branch of f m in some
open disk centered at z0 .
If the range of f contained in a region with a branch l of log z, then f ◦ l
is a branch of log f , then f ◦ l is a branch of log f .

f (z) = (z − z0 )m g(z)
g(z0 ) ̸= so in some neighborhood of z0 , g(z) ̸= 0.
but, (z − z0 )m = 0 at z0 so, how can there be a branch of log?

21
Exercise VII.14.1
Prove that
 there is nno holomorphic function f in the open unit disk such
1 (−1)
that f = for n = 2, 3, . . ..
n n2
Suppose such an f . z 2 is the same as f for a subsequence that converges
to 0. But f ̸= z 2 .

22
Exercise VII.14.2
(a)
Let G be a nonempty, connected, open subset of C which is symmetric with
respect to the real axis. Let f be a holomorphic function in G such that f is
real valued on G ∩ R. Prove that f (z) = f (z) for all z in G.
There must be a neighborhood B of a real number x ∈ G entirely con-
tained in G since otherwise G would be empty or disconnected.
Let (xn ) be a sequence of real numbers entirely contained in B that
converges to x.
In exercise 2.8.2 it was shown that the function if f is holomorphic, then
g(z) = f (z) is holomorphic. Notice that f (xn ) − g(xn ) is 0 for all n and
converges to 0. So, f (z) − g(z) is 0 for all z ∈ G, meaning f (z) = f (z).

(b)
Let G be as in part (a), and let f be a holomorphic function G such that f is
real valued on a nonempty subinterval of G ∩ R. Prove that f is real-valued
on all of G ∩ R.
With f as in the problem statement we can define for z ∈ G, g(z) =
f (z) which is holomorphic as shown in exercise 2.8.2. Essentially the same
argument as above can be used to conclude that f = g except that the
sequence is chosen in the nonempty subinterval of G ∩ R.
Now for any x ∈ G ∩ R we have f (x) = f (x), meaning f (x) is real.

23