I am wondering about the following, possibly naive, question.

Suppose I have a smooth curve, which intersects the horizontal axis in a manner that leads to an accumulation point. More precisely, given a $C^2$ curve $y=f(x)$ such that the set $\{x$ s.t. $f(x)=0\}$ is a set of measure zero (for the Lebesgue measure) which contains an accumulation point, say $x=0$. Does it mean that the $\mathcal{H}^1$ Hausdorff measure of the curve (as a set in $\mathbb{R}^2$) is infinite?

(I hope that it is the case, but I would be very happy with a counter-example as well).


No, it does not have to be infinite. For example, the graph of $y=x^{3}\sin (1/x)$, $-1\le x\le 1$, is $C^2$ smooth and its length is finite.

Actually, every $C^2$ curve with compact parameter domain has finite length.

  • $\begingroup$ Thanks! On your example: does any measure type tool distinguish this graph from y=x (due to the accumulation of zeroes), or am I looking in the wrong direction? $\endgroup$ – walt Mar 24 '14 at 16:28
  • $\begingroup$ @walt Hausdorff measure and dimension are too rough for this; you need finer tools. It is true that if $\Gamma$ crosses the line transversally, and $\tilde \Gamma$ is $C^1$-close to $\Gamma$, then transversality holds also for $\tilde \Gamma$, hence the number of intersections is finite. This does not work in $C^0$ norm. $\endgroup$ – user127096 Mar 24 '14 at 16:32
  • $\begingroup$ That's very helpful, thanks again! $\endgroup$ – walt Mar 24 '14 at 16:34

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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