Circular distributions and Euler systems (Q5935878)

This article is based on the following observation: Euler systems provide bounds on class groups of real cyclotomic fields. Due to the analytic class number formula, the bounds provided by (Euler systems of) cyclotomic units are already best possible, so in a sense to be made precise, every Euler system has already to be cyclotomic. The exact statement is given in Theorem 4.2; equality is only obtained outside a set of ``bad'' primes. In the main step of the proof (Euler systems provide bounds on the classgroup size, by the machinery of Kolyvagin and Rubin), a precise reference would have been useful to the reader. In the list of references we find Rubin's important article [11]; however it does not literally cover what is needed. In the final section, the author proves a sharper equality statement (no bad primes), assuming that the Greenberg conjecture is valid. There are misprints, some of them confusing, for example: the conductor \(t\) of \(F\) is written \(f\) in the first display on page 373 and the proof of Lemma 4.1, and \(N\) in the statement of Lemma 4.1. Final remark: on page 375 one may replace ``because it follows from Leopoldt's conjecture'' by ``because it is equivalent to Leopoldt's conjecture''.