A note on perinormal domains (Q1622002)

Let \(D\) be an integral domain with quotient field \(K\). An overring of \(D\) means a subring of \(K\) containing \(D\), and \(D\) is said to be perinormal if every local going down overring of \(D\) is a localization of \(D\). The notion of perinormal domains was introduced in [\textit{N. Epstein} and \textit{J. Shapiro}, J. Algebra 451, 65--84 (2016; Zbl 1346.13039)]. Then, among other things, Epstein and Shapiro showed that \(D\) is perinormal if and only if every going down overring of \(D\) is flat over \(D\); a generalized Krull domain is perinormal; and if \(D\) is perinormal, then \(D_P\) is a valuation domain for all height-one prime ideals \(P\) of \(D\). Recall that \(D\) is a P-domain if \(D_P\) is a valuation domain for each prime ideal \(P\) of \(D\) which is minimal over \(aD:b\) for some \(a, b \in D\). Hence, generalized Krull domain \(\Rightarrow\) Prüfer \(v\)-multiplication domain (P\(v\)MD) \(\Rightarrow\) P-domain. In this paper under reivew, the authors proved that every P-domain is perinormal; a P\(v\)MD is just a \(w\)-treed perinormal domain; and if \(D\) is a P\(v\)MD with torsion class group, then \(D\) is globally perinormal (i.e., every going down overring of \(D\) is a fraction ring of \(D\)). They also constructed perinormal and non-perinormal domains by using the technique of pullbacks.