Three more proofs of two congruences for Merca's partition function (Q6876555)

In this brief note, the author presents three new and remarkably concise proofs of two modulo \(2\) congruences for Merca's partition function \(a(n)\), which counts partitions in which odd parts appear with multiplicity at most two (equivalently, partitions with no parts congruent to \(3 \pmod{6}\)). The congruences, namely that \(a(4m+2)\) and \(a(4m+3)\) are identically even, were originally established by \textit{M. Merca} [Aequationes Math. 99, No. 3, 1153--1173 (2025; Zbl 08060173)].\N\NThe main contribution of the paper lies in the extreme brevity and clarity of the new arguments. Each proof relies on simple yet effective manipulations of generating functions modulo \(2\), together with classical identities for Euler products and well-known parity results for \(t\)-core partitions and eta-quotients. In particular, the author highlights how the congruences follow naturally from identities involving \(f_t(q)=\prod_{n\ge1}(1-q^{tn})\) and from structural properties of binary \(q\)-series.\N\NBeyond reproving the known congruences, the paper places the function \(a(n)\) in the broader framework of eta-quotients. This perspective leads to compelling conjectural statements on the global parity behavior of \(a(n)\), supported by extensive numerical evidence. Most notably, it is conjectured that \(a(n)\) is odd with natural density \(1/4\), and that, apart from the progressions \(4m+2\) and \(4m+3\), there are no arithmetic progressions on which \(a(n)\) is identically even.\N\NThe exposition is concise, accessible, and well motivated, making the note a valuable addition to the literature on parity questions in partition theory. It not only refines existing results with sharper proofs, but also opens clear directions for future work on the distribution of coefficients of eta-quotients modulo \(2\).