I use dimH(E) to denote the Hausdorff dimension of a set EsubseteqmathbbR and |E| to denote its Lebesgue measure. It is easy to see from the definition of Hausdorff dimension that if dimH(E)<1, then |E|=0. The converse is not true, and there are many cases where dimH(E)=1 yet |E|=0. So the question:
What was the first (or most elementary) example of this phenomenon?
After some looking around, I was able to prove that a central Cantor set C with ratio of dissection rk=1/(2+frac1k) satisfies the condition I want. It is easy to see that |C|=0 since at step n of the process to construct this Cantor set, it has measure 2n(r1cdotsrn) which in this case limits to 0, but for the Hausdorff dimension I required a non-trivial result from the paper Sums of Cantor sets (Cabrelli, Hare, Molter) that gave the formula
dimH(C)=liminfnfracnln2lnr1cdotsrn.
This result is fairly recent and sophisticated, and I feel that there should be older and simpler examples.
No comments:
Post a Comment