Notes on Complex Analysis
Export PDFGitHub

§4.4 Further Properties of Meromorphic and Entire Functions

4.4 Further Properties of Meromorphic and Entire Functions

Theorem 4.4.1.

Let 𝑈⊆ℂ be a region and 𝑓:𝑈→ℂ be meromorphic. Let 𝛾⊂𝑈 be a positively oriented Jordan curve that is null-homotopic in 𝑈. If 𝑓 has no zeros on 𝛾, then 𝑓 has finitely many zeros and poles in the region bounded by 𝛾. Denote the zeros of 𝑓 in the bounded region by 𝑎1,…,𝑎𝑘 with respective multiplicities 𝛼1,…,𝛼𝑘, and the poles by 𝑏1,…,𝑏𝑚 with respective orders 𝛽1,…,𝛽𝑚. Let 𝜓 be any function holomorphic on a neighborhood of the closure of the bounded region. Then

12πi∮𝛾𝜓(𝑧)𝑓′(𝑧)𝑓(𝑧)d𝑧=∑𝑖=1𝑘𝛼𝑖𝜓(𝑎𝑖)−∑𝑗=1𝑚𝛽𝑗𝜓(𝑏𝑗).
Proof.

Choose disks 𝐷(𝑎𝑖,𝜀𝑖) with pairwise disjoint closures around each zero 𝑎𝑖 and 𝐷(𝑏𝑗,𝜀𝑗′) around each pole 𝑏𝑗, with 𝜀𝑖,𝜀𝑗′>0 sufficiently small so that these disks are contained in int(𝛾), disjoint from 𝛾, and contained in the neighborhood where 𝜓 is holomorphic. The function

𝑔(𝑧)=𝜓(𝑧)𝑓′(𝑧)𝑓(𝑧)

is holomorphic on

int(𝛾)∖(⋃𝑖=1𝑘𝐷(𝑎𝑖,𝜀𝑖)∪⋃𝑗=1𝑚𝐷(𝑏𝑗,𝜀𝑗′)),

since 𝜓 is holomorphic there, 𝑓 is meromorphic with no other singularities, and 𝑓≠0 on 𝛾. The oriented boundary of this punctured domain is 𝛾+∪⋃𝑖=1𝑘𝜕𝐷(𝑎𝑖,𝜀𝑖)−∪⋃𝑗=1𝑚𝜕𝐷(𝑏𝑗,𝜀𝑗′)−. By Cauchy–Goursat (Theorem 3.1.9Theorem 3.1.9Theorem 3.1.9),

∮𝛾+𝑔(𝑧)d𝑧+∑𝑖=1𝑘∮𝜕𝐷(𝑎𝑖,𝜀𝑖)−𝑔(𝑧)d𝑧+∑𝑗=1𝑚∮𝜕𝐷(𝑏𝑗,𝜀𝑗′)−𝑔(𝑧)d𝑧=0.

Thus,

−∮𝛾𝑔(𝑧)d𝑧=∑𝑖=1𝑘∮𝜕𝐷(𝑎𝑖,𝜀𝑖)+𝑔(𝑧)d𝑧+∑𝑗=1𝑚∮𝜕𝐷(𝑏𝑗,𝜀𝑗′)+𝑔(𝑧)d𝑧.

Near each zero 𝑎𝑖, write 𝑓(𝑧)=(𝑧−𝑎𝑖)𝛼𝑖ℎ(𝑧) where ℎ is holomorphic at 𝑎𝑖 with ℎ(𝑎𝑖)≠0. Then

𝑓′(𝑧)𝑓(𝑧)=𝛼𝑖𝑧−𝑎𝑖+ℎ′(𝑧)ℎ(𝑧),

so

𝑔(𝑧)=𝜓(𝑧)(𝛼𝑖𝑧−𝑎𝑖+ℎ′(𝑧)ℎ(𝑧)).

Then,

∮𝜕𝐷(𝑎𝑖,𝜀𝑖)𝑔(𝑧)d𝑧=∮𝜕𝐷(𝑎𝑖,𝜀𝑖)𝜓(𝑧)(𝛼𝑖𝑧−𝑎𝑖+ℎ′(𝑧)ℎ(𝑧))d𝑧=2πi𝛼𝑖𝜓(𝑎𝑖),

where the first term has been reduced by the Cauchy–Goursat Formula (Theorem 3.1.10Theorem 3.1.10Theorem 3.1.10) and the second integral vanishes by the Cauchy–Goursat Theorem (Theorem 3.1.9Theorem 3.1.9Theorem 3.1.9).

Near a pole 𝑏𝑗, write 𝑓(𝑧)=(𝑧−𝑏𝑗)−𝛽𝑗𝑘(𝑧) where 𝑘 is holomorphic at 𝑏𝑗 with 𝑘(𝑏𝑗)≠0. Then

𝑓′(𝑧)𝑓(𝑧)=−𝛽𝑗𝑧−𝑏𝑗+𝑘′(𝑧)𝑘(𝑧),

so

𝑔(𝑧)=𝜓(𝑧)(−𝛽𝑗𝑧−𝑏𝑗+𝑘′(𝑧)𝑘(𝑧)).

A similar calculation yields that

∮𝜕𝐷(𝑏𝑗,𝜀𝑗′)𝑔(𝑧)d𝑧=−2πi𝛽𝑗𝜓(𝑏𝑗).

Combining these,

∮𝛾𝜓(𝑧)𝑓′(𝑧)𝑓(𝑧)d𝑧=∑𝑖=1𝑘2πi𝛼𝑖𝜓(𝑎𝑖)−∑𝑗=1𝑚2πi𝛽𝑗𝜓(𝑏𝑗)=2πi(∑𝑖=1𝑘𝛼𝑖𝜓(𝑎𝑖)−∑𝑗=1𝑚𝛽𝑗𝜓(𝑏𝑗)).□

Theorem 4.4.2 (Argument Principle).

Let 𝑈⊆ℂ be a region and 𝑓:𝑈→ℂ be meromorphic. Let 𝛾⊂𝑈 be a simple, closed, positively oriented curve that is null-homotopic in 𝑈. If 𝑓 has no zeros or poles on 𝛾, then 𝑓 has finitely many zeros and poles in the region bounded by 𝛾, and the number of zeros, 𝑘, minus the number of poles, 𝑘′, counting multiplicities and orders, is given by

𝑘−𝑘′=12πi∮𝛾𝑓′(𝑧)𝑓(𝑧)d𝑧.
Let Γ be the image of 𝛾 under the map 𝑤=𝑓(𝑧). Then 𝑘−𝑘′=IndΓ(0).
Proof.

By Theorem 4.4.1Theorem 4.4.1Theorem 4.4.1 for 𝜓≡1,

12πi∮𝛾𝑓′(𝑧)𝑓(𝑧)d𝑧=𝑘−𝑘′.

Parametrize Γ by 𝑤=𝑓(𝑧). Then d𝑤=𝑓′(𝑧)d𝑧, and

𝑘−𝑘′=12πi∮Γd𝑤𝑤=IndΓ(0).□