Let’s say the functionf is piecewise continuous on the integration pathC:z=z(t),t∈[a,b]. If the positive number L=∫ab∣z’(t)∣dt is the length of C, and for all points on C there exists a positive number M satisfying ∣f(z)∣≤M, then
∫Cf(z)dz≤ML
Proof
For the function z’:[a,b]→C let’s say ∫abz’(t)dt=r.
If r=0 then it can be represented as ∫abz’(t)dt=reiθ. Then θ is a constant, therefore
r=∫abe−iθz’(t)dt≤∫abe−iθz’(t)dt=∫abe−iθ∣z’(t)∣dt
Here, because the magnitude of a complex number’s real root is always 1, e−iθ=1. Thus,
∫abz’(t)dt=r≤∫ab∣z’(t)∣dt
This demonstrates that the properties of definite integrals which hold for real functions are also valid for complex functions. Using the inequality derived above, we get
∫Cf(z)dz=≤==∫abf(z(t))z’(t)dt∫ab∣f(z(t))∣∣z’(t)∣dt∫abM∣z’(t)∣dtML
■
Explanation
In the ML Lemma, M stands for Maximum, and L stands for Length.
What can be confusing when using the ML Lemma is how to choose L. Often in the original integral, C is given as a circle with radius r, and when substituting, the integration interval becomes [0,2π]. After substituting and applying the ML Lemma, one should use 2π, the length of the new integration interval, not 2πr, the original circle’s circumference. In other words, it’s important to remember that you must apply the ML Lemma to the new curve (segment), C′, that emerges through substitution.
Osborne (1999). Complex variables and their applications: p76. ↩︎