Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. It's 100% free, no registration required.

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

I am having trouble with the following problem and would appreciate some help. It is a piece of a proof of Cauchy's integral formula for a circle that appears in Dieudonne's Treatise on Analysis, Vol. I. I am interested in this proof because it proceeds directly from the assumption that $f$ is complex differentiable in an open neighborhood of the disk. All other proofs that I have seen take a more indirect route, often going through some form of Cauchy's theorem first.

Let $f$ be complex differentiable on an open neighbourhood of the closed unit disc. Assume that, for each given $\lambda \in (0,1)$ and $\epsilon > 0$, there are points $t_0 = 0 < t_1 < \cdots < t_m = 2\pi$, a number $\delta > 0$, and in each interval $[t_k,t_{k+1}]$ a point $\theta_k$ such that $$ | f(\zeta_k + x) - f(\zeta_k) - f^{\prime}(\zeta_k)x | \leq \epsilon |x| $$ whenever $|h| \leq \delta$ and $t_k \leq t \leq t_{k+1}$. Here $\zeta_k = z + \lambda(e^{i\theta_k} - z)$ and $\zeta_k + x = z + (\lambda+h)(e^{it} - z)$.

Define $$ g(\lambda) = \int_{0}^{2\pi} \frac{ f(z + \lambda(e^{it} - z)) - f(z)}{e^{it} - z} e^{it} dt. $$

By comparing each integral $$ A_k = \int_{t_k}^{t_{k+1}} \frac{ f( z + (\lambda + h)(e^{it} - z) ) - f( z + \lambda(e^{it} - z) ) }{e^{it} - z} e^{it} dt $$ to the corresponding expression $$ B_k = \frac{h}{i \lambda} ( f( z + \lambda(e^{i t_{k+1}} - z) ) - f( z + \lambda(e^{i t_k} - z) ) ) $$ show that, for each given $\lambda \in (0,1)$ and $\epsilon > 0$, there is a $\delta > 0$ such that $$ | g(\lambda + h) - g(\lambda) - 0 \cdot h | \leq \epsilon |h| $$ for $|h| \leq \delta$. That is, show that $g^{\prime}(\lambda) = 0$ for $\lambda \in (0,1)$.

Note $$ \sum_{k=0}^{m-1} (A_k - B_k) = g(\lambda + h) - g(\lambda) - 0 \cdot h. $$

share|cite|improve this question
Are you sure that the $\epsilon$ at the end is the same $\epsilon$ as in the premise and not some multiple of it? (I looked at Problem 1 in Section IX.10 of the book; this seems to be the problem you're referring to, and to solve that problem it would suffice if the $\epsilon$ in your conclusion were some fixed multiple of the one in your premise, and that might be easier to prove.) – joriki Mar 2 '11 at 3:30
That's true. The $\epsilon$ in the conclusion and the $\epsilon$ in the premise don't need to be the same. I was already aware of this, but I appreciate your suggestion. I have edited the question slightly to make it more clear that the $\epsilon$'s need not be the same. – admchrch Mar 2 '11 at 5:10

The idea of the proof sketched in the question is to use conformality to relate the effect of normal perturbation of the circle of integration (which expands it from radius $\lambda$ to $\lambda+h$) to the tangential perturbation of said circle (which does nothing to the integral). Indeed, the complex differentiability gives $$\frac{ f( z + (\lambda + h)(e^{it} - z) ) - f( z + \lambda(e^{it} - z) ) }{e^{it} - z}e^{it} = he^{it}f'( z + \lambda(e^{it} - z) ) + o(h) \tag1$$ and $$\frac{ f( z + \lambda (e^{i(t+h)} - z) ) - f( z + \lambda(e^{it} - z) ) }{i\lambda } = he^{it}f'( z + \lambda(e^{it} - z) ) + o(h)\tag2$$ where the right-hand sides are the same. Now let $t_k=hk$, so that $t+h$ in (2) becomes $t_{k+1}$ and the sum over $k$ telescopes. The same summation in (1) gives a Riemann sum of $$ \int_0^{2\pi} \frac{ f( z + (\lambda + h)(e^{it} - z) ) - f( z + \lambda(e^{it} - z) ) }{e^{it} - z} e^{it} dt$$ which is thus shown to be $o(h)$.

share|cite|improve this answer
They invented the revival badge for you, +1. – 1015 Jun 23 '13 at 14:05

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.