On the first case of the Fermat theorem for cyclotomic fields (Q5960207)

The author studies the Fermat equation (*) \(x^l+y^l+z^l=0\) assuming that \(x,y,z\) are integers prime to \(l\) in the \(l\)th cyclotomic field (\(l\) a prime \(>3\)). He generalizes classical criteria proved in the case of rational integers \(x,y,z\) by Kummer, Mirimanoff and Vandiver. As corollaries he obtains several smooth criteria. He may be unaware of some previous work done on this topic. For example, \textit{T. Morishima} [Jap. J. Math. 11, 241--252 (1935; Zbl 0011.33804); reproduced on pp. 171--182 in Collected papers, Queen's Papers in Pure and Applied Math. 84, Queen's Univ., Kingston, ON (1990; Zbl 0746.11002)] proved that the solvability of (*) under the conditions in question would imply that \(q^l \equiv q \pmod {l^2}\) for all primes \(q \leq 31\). This is essentially the author's theorem 4, with 31 replaced by 89 (corresponding to the present status in computational results). The author's method is quite different, however. Among his corollaries there are also some observations relating the solvability to the abc-conjecture and to a bound for \(ii(l)\), the index of irregularity of \(l\).