0% found this document useful (0 votes)
96 views28 pages

Complex Analysis: Chapter VI. The Maximum Modulus Theorem VI.4. Phragm En-Lindel of Theorem-Proofs of Theorems

The document summarizes theorems and corollaries related to the Phragmén-Lindelöf theorem from complex analysis. Theorem VI.4.A proves that if an entire function f satisfies |f(z)| ≤ M + |z|^α for 0 < α < 1, then f must be constant. The Phragmén-Lindelöf theorem (Theorem VI.4.1) states that if an analytic function f satisfies certain boundary conditions on a simply connected region G, then |f(z)| is bounded on G. Corollary VI.4.2 gives conditions under which the Phragmén-Lindelöf theorem applies to a region defined by an angular domain

Uploaded by

TOM DAVIS
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
96 views28 pages

Complex Analysis: Chapter VI. The Maximum Modulus Theorem VI.4. Phragm En-Lindel of Theorem-Proofs of Theorems

The document summarizes theorems and corollaries related to the Phragmén-Lindelöf theorem from complex analysis. Theorem VI.4.A proves that if an entire function f satisfies |f(z)| ≤ M + |z|^α for 0 < α < 1, then f must be constant. The Phragmén-Lindelöf theorem (Theorem VI.4.1) states that if an analytic function f satisfies certain boundary conditions on a simply connected region G, then |f(z)| is bounded on G. Corollary VI.4.2 gives conditions under which the Phragmén-Lindelöf theorem applies to a region defined by an angular domain

Uploaded by

TOM DAVIS
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 28

Complex Analysis

Chapter VI. The Maximum Modulus Theorem


VI.4. Phragmén-Lindelöf Theorem—Proofs of Theorems

September 22, 2017

() Complex Analysis September 22, 2017 1 / 13


Table of contents

1 Theorem VI.4.A

2 Theorem VI.4.1. Phragmén-Lindelöf Theorem

3 Corollary VI.4.2

4 Corollary VI.4.4

() Complex Analysis September 22, 2017 2 / 13


Theorem VI.4.A

Theorem VI.4.A
Theorem VI.4.A. Suppose f is an entire function, M > 0, and
0 < α < 1. Suppose |f (z)| ≤ M + |z|α for all z ∈ C. Then f is constant.
Z
0 n! f (w )
Proof. We take n = 1 in Corollary VI.2.13: f (a) = dw
2πi γ (w − a)2
where γ(t) = a + re it and 0 ≤ t ≤ 2π.

() Complex Analysis September 22, 2017 3 / 13


Theorem VI.4.A

Theorem VI.4.A
Theorem VI.4.A. Suppose f is an entire function, M > 0, and
0 < α < 1. Suppose |f (z)| ≤ M + |z|α for all z ∈ C. Then f is constant.
Z
0 n! f (w )
Proof. We take n = 1 in Corollary VI.2.13: f (a) = dw
2πi γ (w − a)2
where γ(t) = a + re it and 0 ≤ t ≤ 2π. We have

|f (w )|
Z Z
0
n! f (w ) n!
|f (a)| =
2
dw ≤
|dw |
2πi γ (w − a) 2π γ |w − a|2
|f (w )|
Z Z
n! n!
= |dw | ≤ (M + |z|α ) |dw |
2π γ r 2 2πr 2 γ
Z
n!
≤ (M + (|a| + r )α ) |dw | since |w | ≤ |a| + r
2πr 2 γ
n!
= 2πr (M + (|a| + r )α ).
2πr 2

() Complex Analysis September 22, 2017 3 / 13


Theorem VI.4.A

Theorem VI.4.A
Theorem VI.4.A. Suppose f is an entire function, M > 0, and
0 < α < 1. Suppose |f (z)| ≤ M + |z|α for all z ∈ C. Then f is constant.
Z
0 n! f (w )
Proof. We take n = 1 in Corollary VI.2.13: f (a) = dw
2πi γ (w − a)2
where γ(t) = a + re it and 0 ≤ t ≤ 2π. We have

|f (w )|
Z Z
0
n! f (w ) n!
|f (a)| =
2
dw ≤
|dw |
2πi γ (w − a) 2π γ |w − a|2
|f (w )|
Z Z
n! n!
= |dw | ≤ (M + |z|α ) |dw |
2π γ r 2 2πr 2 γ
Z
n!
≤ (M + (|a| + r )α ) |dw | since |w | ≤ |a| + r
2πr 2 γ
n!
= 2πr (M + (|a| + r )α ).
2πr 2

() Complex Analysis September 22, 2017 3 / 13


Theorem VI.4.A

Theorem VI.4.A (continued)

Theorem VI.4.A. Suppose f is an entire function, M > 0, and


0 < α < 1. Suppose |f (z)| ≤ M + |z|α for all z ∈ C. Then f is constant.

Proof (continued). . . .

n!
|f 0 (a)| ≤ 2πr (M + (|a| + r )α ).
2πr 2
Since this holds for arbitrary r (f is entire), then we see that r → ∞
implies that f 0 (a) = 0. Also, a is arbitrary, so f 0 (z) = 0 for all z ∈ C and
hence f is constant.

() Complex Analysis September 22, 2017 4 / 13


Theorem VI.4.1. Phragmén-Lindelöf Theorem

Proposition VI.4.1
Theorem VI.4.1. Phragmén-Lindelöf Theorem.
Let G be a simply connected region and let f be an analytic function on
G . Suppose there is an analytic function φ : G → C which is nonzero and
is bounded on G . If M is a constant and ∂∞ G = A ∪ B such that
(a) for every a ∈ A we have lim supz→a |f (z)| ≤ M, and
(b) for every b ∈ B and η > 0, we have
lim supz→b |f (z)||φ(z)|η ≤ M,
then |f (z)| ≤ M for all z ∈ G .
Proof. Let |ϕ(z)| ≤ κ for all z ∈ G . Since G is simply connected and ϕ is
nonzero on G , then by Corollary IV.6.17, there is a branch of log ϕ(z) on
G . Hence g (z) = exp(η log ϕ(z)) is an analytic branch of (ϕ(z))η for
η > 0, and |g (z)| = |ϕ(z)|η . Define F : G → C as F (z) = f (z)g (z)κ−η .
Then F is analytic on G and |F (z)| ≤ |f (z)|. Now for a ∈ ∂∞ G for which
condition (a) holds, we have lim supz→a |F (z)| ≤ lim supz→∞ |f (z)| ≤ M.
() Complex Analysis September 22, 2017 5 / 13
Theorem VI.4.1. Phragmén-Lindelöf Theorem

Proposition VI.4.1
Theorem VI.4.1. Phragmén-Lindelöf Theorem.
Let G be a simply connected region and let f be an analytic function on
G . Suppose there is an analytic function φ : G → C which is nonzero and
is bounded on G . If M is a constant and ∂∞ G = A ∪ B such that
(a) for every a ∈ A we have lim supz→a |f (z)| ≤ M, and
(b) for every b ∈ B and η > 0, we have
lim supz→b |f (z)||φ(z)|η ≤ M,
then |f (z)| ≤ M for all z ∈ G .
Proof. Let |ϕ(z)| ≤ κ for all z ∈ G . Since G is simply connected and ϕ is
nonzero on G , then by Corollary IV.6.17, there is a branch of log ϕ(z) on
G . Hence g (z) = exp(η log ϕ(z)) is an analytic branch of (ϕ(z))η for
η > 0, and |g (z)| = |ϕ(z)|η . Define F : G → C as F (z) = f (z)g (z)κ−η .
Then F is analytic on G and |F (z)| ≤ |f (z)|. Now for a ∈ ∂∞ G for which
condition (a) holds, we have lim supz→a |F (z)| ≤ lim supz→∞ |f (z)| ≤ M.
() Complex Analysis September 22, 2017 5 / 13
Theorem VI.4.1. Phragmén-Lindelöf Theorem

Proposition VI.4.1

Proof. For b ∈ ∂∞ G for which condition (b) holds, we have

lim sup |F (z)| = lim sup |f (z)g (z)κ−η | = κ−η lim sup |f (z)||ϕ(z)|η ≤ κ−η M.
z→b z→b z→b

So F satisfies the hypotheses of the Maximum Modulus Theorem—Third


Version (Theorem VI.1.4) with M of Theorem VI.1.4 replaced with
max{M, κ−η M} here. so, by Theorem VI.1.4, |f (z)| ≤ max{M, κ−η M} for
all z ∈ G . So

κη

|F (z)| κ
|f (z)| = κη = |F (z)| ≤ max{M, κ−η M}
|g (z)| |ϕ(z)|η |ϕ(z)|

for all z ∈ G and for all η > 0. Letting η → 0+ implies |f (z)| ≤ M for all
z ∈ G.

() Complex Analysis September 22, 2017 6 / 13


Corollary VI.4.2

Corollary VI.4.2

Corollary VI.4.2. Let a ≥ 1/2 and let G = {z | |arg(z)| < π/(2a)}.


Suppose that f is analytic on G and suppose there is a constant M such
that lim supz→w |f (z)| ≤ M for all w ∈ ∂G . If there are positive constants
P and b < a such that |f (z)| ≤ P exp(|z|b ) for all z with |z| sufficiently
large, then |f (z)| ≤ M for all z ∈ G .
Proof. Let 0 < b < c < a and define ϕ(z) = exp(−z c ) for z ∈ G . If
z = re iθ where |θ| < π/(2a), then Re(z c ) = r c cos(cθ). So for z ∈ G ,

|ϕ(z)| = | exp(−z c )| = | exp(Re(−z c ))| = exp(−r c cos(cθ))

when z = re iθ .

() Complex Analysis September 22, 2017 7 / 13


Corollary VI.4.2

Corollary VI.4.2

Corollary VI.4.2. Let a ≥ 1/2 and let G = {z | |arg(z)| < π/(2a)}.


Suppose that f is analytic on G and suppose there is a constant M such
that lim supz→w |f (z)| ≤ M for all w ∈ ∂G . If there are positive constants
P and b < a such that |f (z)| ≤ P exp(|z|b ) for all z with |z| sufficiently
large, then |f (z)| ≤ M for all z ∈ G .
Proof. Let 0 < b < c < a and define ϕ(z) = exp(−z c ) for z ∈ G . If
z = re iθ where |θ| < π/(2a), then Re(z c ) = r c cos(cθ). So for z ∈ G ,

|ϕ(z)| = | exp(−z c )| = | exp(Re(−z c ))| = exp(−r c cos(cθ))

when z = re iθ . Since c < a, cos(cθ) ≥ ρ > 0 for some ρ (since c < a


implies cθ < aθ < a(π/(2a)) = π/2 for z ∈ G ). So
|ϕ(z)| = | exp(−z c )| = exp(−r c cos(cθ)) ≤ exp(−r c ρ) for all z ∈ G , and
ϕ is bounded on G . Also, if η > 0 and z = re iθ is sufficiently large, then

() Complex Analysis September 22, 2017 7 / 13


Corollary VI.4.2

Corollary VI.4.2

Corollary VI.4.2. Let a ≥ 1/2 and let G = {z | |arg(z)| < π/(2a)}.


Suppose that f is analytic on G and suppose there is a constant M such
that lim supz→w |f (z)| ≤ M for all w ∈ ∂G . If there are positive constants
P and b < a such that |f (z)| ≤ P exp(|z|b ) for all z with |z| sufficiently
large, then |f (z)| ≤ M for all z ∈ G .
Proof. Let 0 < b < c < a and define ϕ(z) = exp(−z c ) for z ∈ G . If
z = re iθ where |θ| < π/(2a), then Re(z c ) = r c cos(cθ). So for z ∈ G ,

|ϕ(z)| = | exp(−z c )| = | exp(Re(−z c ))| = exp(−r c cos(cθ))

when z = re iθ . Since c < a, cos(cθ) ≥ ρ > 0 for some ρ (since c < a


implies cθ < aθ < a(π/(2a)) = π/2 for z ∈ G ). So
|ϕ(z)| = | exp(−z c )| = exp(−r c cos(cθ)) ≤ exp(−r c ρ) for all z ∈ G , and
ϕ is bounded on G . Also, if η > 0 and z = re iθ is sufficiently large, then

() Complex Analysis September 22, 2017 7 / 13


Corollary VI.4.2

Corollary VI.4.2 (continued)

Proof (continued).

|f (z)||ϕ(z)|η ≤ P exp(|z|b )|ϕ(z)|η by hypothesis


≤ P exp(|z|b )(exp(−r c ρ))η
= P exp(r b − ηr c ρ) since z = re iθ

(the “sufficiently large” is required of z = re iθ to get the bound


|f (z)| ≤ P exp(|z|b ) for z “sufficiently large” as is hypothesized). But
r b − ηr c ρ = r c (r b−c − ηρ). Since b < c, we have b − c < 0 and so
r b−c → 0+ as r → +∞. so we have r b − ηr c ρ − r c (r b−c − ηρ) → −∞ as
r → +∞. Now lim supz→a |f (z)| ≤ M for all a ∈ ∂G by hypothesis. For
b = ∞, lim supz → ∞||f (z)||ϕ(z)|η = limr →∞ P exp(r b − ηr c ρ) = 0 ≤ M.

() Complex Analysis September 22, 2017 8 / 13


Corollary VI.4.2

Corollary VI.4.2 (continued)

Proof (continued).

|f (z)||ϕ(z)|η ≤ P exp(|z|b )|ϕ(z)|η by hypothesis


≤ P exp(|z|b )(exp(−r c ρ))η
= P exp(r b − ηr c ρ) since z = re iθ

(the “sufficiently large” is required of z = re iθ to get the bound


|f (z)| ≤ P exp(|z|b ) for z “sufficiently large” as is hypothesized). But
r b − ηr c ρ = r c (r b−c − ηρ). Since b < c, we have b − c < 0 and so
r b−c → 0+ as r → +∞. so we have r b − ηr c ρ − r c (r b−c − ηρ) → −∞ as
r → +∞. Now lim supz→a |f (z)| ≤ M for all a ∈ ∂G by hypothesis. For
b = ∞, lim supz → ∞||f (z)||ϕ(z)|η = limr →∞ P exp(r b − ηr c ρ) = 0 ≤ M.
So f satisfies the hypotheses of the Phragmén-Lindelöf Theorem (as does
nonzero, analytic ϕ), and so |f (z)| ≤ M for z ∈ G .

() Complex Analysis September 22, 2017 8 / 13


Corollary VI.4.2

Corollary VI.4.2 (continued)

Proof (continued).

|f (z)||ϕ(z)|η ≤ P exp(|z|b )|ϕ(z)|η by hypothesis


≤ P exp(|z|b )(exp(−r c ρ))η
= P exp(r b − ηr c ρ) since z = re iθ

(the “sufficiently large” is required of z = re iθ to get the bound


|f (z)| ≤ P exp(|z|b ) for z “sufficiently large” as is hypothesized). But
r b − ηr c ρ = r c (r b−c − ηρ). Since b < c, we have b − c < 0 and so
r b−c → 0+ as r → +∞. so we have r b − ηr c ρ − r c (r b−c − ηρ) → −∞ as
r → +∞. Now lim supz→a |f (z)| ≤ M for all a ∈ ∂G by hypothesis. For
b = ∞, lim supz → ∞||f (z)||ϕ(z)|η = limr →∞ P exp(r b − ηr c ρ) = 0 ≤ M.
So f satisfies the hypotheses of the Phragmén-Lindelöf Theorem (as does
nonzero, analytic ϕ), and so |f (z)| ≤ M for z ∈ G .

() Complex Analysis September 22, 2017 8 / 13


Corollary VI.4.4

Corollary VI.4.4
Corollary VI.4.4. Let a ≥ 1/2 and let G = {z | arg(z) < π/(2a)}, and
suppose that for every w ∈ ∂G , lim supz→w |f (z)| ≤ M. Moreover, assume
that for every δ > 0 there is a constant P (which may depend on δ) such
that |f (z)| ≤ P exp(δ|z|a ) for z ∈ G and |z| sufficiently large. Then
|f (z)| ≤ M for all z ∈ G .
Proof. Define F : G → C as F (z) = f (z) exp(−εz 1 ) where ε > 0 is fixed.
If x > 0 and δ satisfies 0 < δ < ε then there is a constant P with
|f (x)| = |f (x) exp(−εx a )|
≤ P exp(δx a ) exp(−εx a ) for x sufficiently large
= P exp((δ − ε)x a ).

() Complex Analysis September 22, 2017 9 / 13


Corollary VI.4.4

Corollary VI.4.4
Corollary VI.4.4. Let a ≥ 1/2 and let G = {z | arg(z) < π/(2a)}, and
suppose that for every w ∈ ∂G , lim supz→w |f (z)| ≤ M. Moreover, assume
that for every δ > 0 there is a constant P (which may depend on δ) such
that |f (z)| ≤ P exp(δ|z|a ) for z ∈ G and |z| sufficiently large. Then
|f (z)| ≤ M for all z ∈ G .
Proof. Define F : G → C as F (z) = f (z) exp(−εz 1 ) where ε > 0 is fixed.
If x > 0 and δ satisfies 0 < δ < ε then there is a constant P with
|f (x)| = |f (x) exp(−εx a )|
≤ P exp(δx a ) exp(−εx a ) for x sufficiently large
= P exp((δ − ε)x a ).
But then |F (x)| → 0 as x → ∞ (x ∈ R). So
M1 = sup{|F (x)| | 0 < x < ∞} < ∞ (since, say, |F (x)| ≤ 1 for x
sufficiently large and then F is continuous on the complement of
“sufficiently large” and so has a MAX there).
() Complex Analysis September 22, 2017 9 / 13
Corollary VI.4.4

Corollary VI.4.4
Corollary VI.4.4. Let a ≥ 1/2 and let G = {z | arg(z) < π/(2a)}, and
suppose that for every w ∈ ∂G , lim supz→w |f (z)| ≤ M. Moreover, assume
that for every δ > 0 there is a constant P (which may depend on δ) such
that |f (z)| ≤ P exp(δ|z|a ) for z ∈ G and |z| sufficiently large. Then
|f (z)| ≤ M for all z ∈ G .
Proof. Define F : G → C as F (z) = f (z) exp(−εz 1 ) where ε > 0 is fixed.
If x > 0 and δ satisfies 0 < δ < ε then there is a constant P with
|f (x)| = |f (x) exp(−εx a )|
≤ P exp(δx a ) exp(−εx a ) for x sufficiently large
= P exp((δ − ε)x a ).
But then |F (x)| → 0 as x → ∞ (x ∈ R). So
M1 = sup{|F (x)| | 0 < x < ∞} < ∞ (since, say, |F (x)| ≤ 1 for x
sufficiently large and then F is continuous on the complement of
“sufficiently large” and so has a MAX there).
() Complex Analysis September 22, 2017 9 / 13
Corollary VI.4.4

Corollary VI.4.4 (continued 1)


Proof (continued). Define M2 = max{M1 , M} and
H+ = {z ∈ G | 0 < arg(z) < π/(2a)},
H− = {z ∈ G | −π/(2a) < arg(z) < 0}. Notice that H+ and H− are
sectors which share the boundary {x | 0 < x < ∞}. For any
w ∈ ∂H− ∪ ∂H+ with |arg(w )| = π/(2a) we have

lim sup |F (z)| = lim sup |f (z)|| exp(−εz a )|


z→w z→w
= lim sup |f (z)| exp(Re (−εz a ))
z→w
= lim sup |f (z)| exp(−εr a cos(aθ)) for z = re iθ
z→w
≤ lim sup |f (z)| since − εr z cos(aθ) < 0
z→w
because |zθ| < π/2
≤ M by hypothesis (i.e., definition of M).

() Complex Analysis September 22, 2017 10 / 13


Corollary VI.4.4

Corollary VI.4.4 (continued 1)


Proof (continued). Define M2 = max{M1 , M} and
H+ = {z ∈ G | 0 < arg(z) < π/(2a)},
H− = {z ∈ G | −π/(2a) < arg(z) < 0}. Notice that H+ and H− are
sectors which share the boundary {x | 0 < x < ∞}. For any
w ∈ ∂H− ∪ ∂H+ with |arg(w )| = π/(2a) we have

lim sup |F (z)| = lim sup |f (z)|| exp(−εz a )|


z→w z→w
= lim sup |f (z)| exp(Re (−εz a ))
z→w
= lim sup |f (z)| exp(−εr a cos(aθ)) for z = re iθ
z→w
≤ lim sup |f (z)| since − εr z cos(aθ) < 0
z→w
because |zθ| < π/2
≤ M by hypothesis (i.e., definition of M).

() Complex Analysis September 22, 2017 10 / 13


Corollary VI.4.4

Corollary VI.4.4 (continued 2)

Proof (continued). So for all w ∈ ∂H− ∪ ∂H+ we have


lim supz→w |f (z)| ≤ M2 . In addition, we claim that the hypothesized
condition |f (z)| ≤ P exp(δ|z|a ) for all z ∈ G and |z| sufficiently large
implies that F satisfies the hypotheses of Corollary VI.4.2. (We have

|f (z)| = |f (z) exp(−εz a )| ≤ P exp(δ|z|a )| exp(−εz a )|

= P exp(δ|z|a ) exp(Re(−εz a ) = P exp(δ|z|z − εRe(z a ))


but this must be less than or equal to P1 exp(|z|b ) for positive P1 and
0 < b < a?)

() Complex Analysis September 22, 2017 11 / 13


Corollary VI.4.4

Corollary VI.4.4 (continued 2)

Proof (continued). So for all w ∈ ∂H− ∪ ∂H+ we have


lim supz→w |f (z)| ≤ M2 . In addition, we claim that the hypothesized
condition |f (z)| ≤ P exp(δ|z|a ) for all z ∈ G and |z| sufficiently large
implies that F satisfies the hypotheses of Corollary VI.4.2. (We have

|f (z)| = |f (z) exp(−εz a )| ≤ P exp(δ|z|a )| exp(−εz a )|

= P exp(δ|z|a ) exp(Re(−εz a ) = P exp(δ|z|z − εRe(z a ))


but this must be less than or equal to P1 exp(|z|b ) for positive P1 and
0 < b < a?) So. . . applying Corollary VI.4.2 to F (z) gives |F (z)| ≤ M2 for
all z ∈ H+ ∪ H− . So |f (z)| ≤ M2 for all z ∈ G .

() Complex Analysis September 22, 2017 11 / 13


Corollary VI.4.4

Corollary VI.4.4 (continued 2)

Proof (continued). So for all w ∈ ∂H− ∪ ∂H+ we have


lim supz→w |f (z)| ≤ M2 . In addition, we claim that the hypothesized
condition |f (z)| ≤ P exp(δ|z|a ) for all z ∈ G and |z| sufficiently large
implies that F satisfies the hypotheses of Corollary VI.4.2. (We have

|f (z)| = |f (z) exp(−εz a )| ≤ P exp(δ|z|a )| exp(−εz a )|

= P exp(δ|z|a ) exp(Re(−εz a ) = P exp(δ|z|z − εRe(z a ))


but this must be less than or equal to P1 exp(|z|b ) for positive P1 and
0 < b < a?) So. . . applying Corollary VI.4.2 to F (z) gives |F (z)| ≤ M2 for
all z ∈ H+ ∪ H− . So |f (z)| ≤ M2 for all z ∈ G .

() Complex Analysis September 22, 2017 11 / 13


Corollary VI.4.4

Corollary VI.4.4 (continued 3)


Proof (continued). We claim that M2 = M. If not, then M1 = M1 > M.
But then we have that |f (z)| assumes its maximum value in G at some
positive real number x ∈ G because limx→∞ |F (x)| = 0 as argued above
and
lim sup |F (x)| = lim sup |f (x)|| exp(−εx a )|
x→0 x→0
= lim sup |f (x)| exp(0) = lim sup |f (x)| ≤ M < M1 ,
x→0 x→0
so |f | as a continuous function on (0, ∞) must attain its supremum over
this set:

() Complex Analysis September 22, 2017 12 / 13


Corollary VI.4.4

Corollary VI.4.4 (continued 3)


Proof (continued). We claim that M2 = M. If not, then M1 = M1 > M.
But then we have that |f (z)| assumes its maximum value in G at some
positive real number x ∈ G because limx→∞ |F (x)| = 0 as argued above
and
lim sup |F (x)| = lim sup |f (x)|| exp(−εx a )|
x→0 x→0
= lim sup |f (x)| exp(0) = lim sup |f (x)| ≤ M < M1 ,
x→0 x→0
so |f | as a continuous function on (0, ∞) must attain its supremum over
this set:

() Complex Analysis September 22, 2017 12 / 13


Corollary VI.4.4

Corollary VI.4.4 (continued 3)


Proof (continued). We claim that M2 = M. If not, then M1 = M1 > M.
But then we have that |f (z)| assumes its maximum value in G at some
positive real number x ∈ G because limx→∞ |F (x)| = 0 as argued above
and
lim sup |F (x)| = lim sup |f (x)|| exp(−εx a )|
x→0 x→0
= lim sup |f (x)| exp(0) = lim sup |f (x)| ≤ M < M1 ,
x→0 x→0
so |f | as a continuous function on (0, ∞) must attain its supremum over
this set:

() Complex Analysis September 22, 2017 12 / 13


Corollary VI.4.4

Corollary VI.4.4 (continued 4)


Corollary VI.4.4. Let a ≥ 1/2 and let G = {z | arg(z) < π/(2a)}, and
suppose that for every w ∈ ∂G , lim supz→w |f (z)| ≤ M. Moreover, assume
that for every δ > 0 there is a constant P (which may depend on δ) such
that |f (z)| ≤ P exp(δ|z|a ) for z ∈ G and |z| sufficiently large. Then
|f (z)| ≤ M for all z ∈ G .
Proof (continued). But then by the Maximum Modulus Theorem—First
Version (Theorem VI.1.1), f must be a constant and then M = M1 = M2 .
So we have established that M2 = M and |F (z)| ≤ M for all z ∈ G . That
is,

|f (z)| = |F (z) exp(εz a )| by definition of F


≤ M exp(εRe(z a ))

for all z ∈ G . Since M is independent of ε, we can let arbitrary ε → 0 and


conclude that |f (z)| ≤ M for all z ∈ G .
() Complex Analysis September 22, 2017 13 / 13
Corollary VI.4.4

Corollary VI.4.4 (continued 4)


Corollary VI.4.4. Let a ≥ 1/2 and let G = {z | arg(z) < π/(2a)}, and
suppose that for every w ∈ ∂G , lim supz→w |f (z)| ≤ M. Moreover, assume
that for every δ > 0 there is a constant P (which may depend on δ) such
that |f (z)| ≤ P exp(δ|z|a ) for z ∈ G and |z| sufficiently large. Then
|f (z)| ≤ M for all z ∈ G .
Proof (continued). But then by the Maximum Modulus Theorem—First
Version (Theorem VI.1.1), f must be a constant and then M = M1 = M2 .
So we have established that M2 = M and |F (z)| ≤ M for all z ∈ G . That
is,

|f (z)| = |F (z) exp(εz a )| by definition of F


≤ M exp(εRe(z a ))

for all z ∈ G . Since M is independent of ε, we can let arbitrary ε → 0 and


conclude that |f (z)| ≤ M for all z ∈ G .
() Complex Analysis September 22, 2017 13 / 13

You might also like