Two questions on Levesque's cyclotomic unit index (Q910429)

Let n be an integer \(n\not\equiv 2\) (mod 4) and \(\zeta_ n=e^{2\pi i/n}\). Let \({\mathbb{Q}}(\zeta_ n)^+={\mathbb{Q}}(\zeta_ n+\zeta_ n^{- 1})\) denote the maximal real subfield of the cyclotomic field and \(E^+_ n\) denote the unit group of \({\mathbb{Q}}(\zeta_ n)^+\). For any fixed subset D of the set of proper factors of n, \textit{C. Levesque} [Number theory, Proc. Conf. Banff/Alberta (Can.) 1988, 325-338 (1990; see the following review)] introduced a unit system \(C_ n(D)=\{\lambda_ a |\) \(1<a<n/2\), \((a,n)=1\}\) where \(\lambda_ a=\zeta_ n^{b(a)}\prod_{d\in D}(1-\zeta_ n^{ad})(1-\zeta_ n^ d)^{- 1}\), \(b(a)=2^{-1}(1-a)\sum_{d\in D}d\), and obtained the index formula \([E^+_ n:\) \(<\pm C_ n(D)>]=h^+_ n i(D)\) where \(h^+_ n\) is the class number of \({\mathbb{Q}}(\zeta_ n)^+\), \(<\pm C_ n(D)>\) denotes the group generated by -1 and \(C_ n(D)\), the term i(D) is explicitly given. Levesque raised two questions about i(D) in his paper cited above: (A) If \(D^*=D\cup \{d^*\}\) with \(d^* | n\), \(d^*<n\), \(d^*\not\in D\), is \(i(D)<i(D^*)?\) (B) If \(d_ 1\in D_ 1\) and \(d_ 2\in D_ 2\) are such that \(D_ 1\setminus \{d_ 1\}=D_ 2\setminus \{d_ 2\}\) and if \(d_ 1<d_ 2\), is \(i(D_ 2)<i(D_ 1)?\) The author considers these two questions in the cases \(n=p^{\alpha}\) and \(n=p^{\alpha}q^{\beta}\) where p and q are different primes, and shows: (I) Question (A) has affirmative answer for \(n=p^{\alpha}\) and \(n=p^{\alpha}q^{\beta}\). (II) Question (B) has affirmative answer for \(n=p^{\alpha}\) but has negative answer for \(n=p^{\alpha}q^{\beta}\).