Research notes on every problem, and a library of the papers behind them. Built from the open erdos repository.
Updated
Subhajit Ghosh and Koushik Ramachandran, Number of components of polynomial lemniscates: a problem of Erdös, Herzog, and Piranian, J. Math. Anal. Appl. 540 (2024), no. 1, Paper No. 128571 (arXiv:2312.13673v1 of 21 December 2023 is the preprint), answer the problem. For a compact of positive logarithmic capacity (the transfinite diameter), let be the largest number of connected components of over monic of degree with all zeros in . Their Question 1.2 restates the problem of Erdős, Herzog and Piranian: is when , and, when and lies in no closed disc of radius , can equal along a subsequence of degrees? Theorem 2.1(a) gives the first answer: implies , so for all large with a depending on . The paper's notation restricts to positive capacity, so a closed set of capacity lies outside the theorem's statement; it is covered all the same, since is monotone in the set and adding a small closed disc to such a set gives a compact set of capacity . The paper's remark after the theorem shows that is not determined by the capacity. The closed disc of radius and the segment both have capacity . Every lemniscate over the disc has one component, so there, while the segment has . Both sets have , so the remark does not decide the parenthetical variant, whether can be chosen depending only on the capacity. The paper does not treat that variant. Proposition 2.3 gives the second answer: a closed lemniscate , which has capacity , has for all in an infinite set. The proposition holds for every closed lemniscate, including the closed unit disc (), which the question's hypothesis excludes; a lemniscate lying in no closed disc of radius , such as , which contains , meets the hypothesis and answers the question. Theorem 2.4 gives for the closure of a bounded Jordan domain with boundary of capacity . Above capacity , Theorem 2.2 gives for all large under a regularity or connectedness condition. The statements above were read from the arXiv preprint (v1), held on its library card; the proofs were not checked here.
Reviewed. The site's curator, Thomas Bloom, marks Problem 1042 proved and credits this paper in the site's commentary (last edited 12 April 2026).
Refereed. Journal of Mathematical Analysis and Applications 540 (2024), no. 1, Paper No. 128571, as the Crossref record of the DOI gives it.