Arithmetically Cohen-Macaulay curves in \(\mathbb{P}^4\) of degree 4 and genus 0 (Q1360931)

A curve \(C \subset {\mathbb P}^n\) is arithmetically Cohen-Macaulay (ACM) if its homogeneous coordinate ring is a Cohen-Macaulay ring. This implies that \(C\) is locally Cohen-Macaulay and connected. An interesting object of study is the scheme \(\text{Hilb}^{nm+1}({\mathbb P}^n)\), the Hilbert scheme parameterizing subschemes of \({\mathbb P}^n\) with Hilbert polynomial \(nm+1\). It is known that the rational normal curves correspond to the points of a smooth, irreducible open subscheme of \(\text{Hilb}^{nm+1}({\mathbb P}^n)\) of dimension \((n-1)(n+3)\). For \(n=3\) the scheme \(\text{Hilb}^{3m+1}({\mathbb P}^3)\) was studied by \textit{R. Piene} and \textit{M. Schlessinger} [Am. J. Math. 107, 761-774 (1985; Zbl 0589.14009)], where it was shown among other things that \(\text{Hilb}^{3m+1}({\mathbb P}^3)\) consists of two smooth components, intersecting transversally, and that every such locally Cohen-Macaulay curve is ACM. For \(n \geq 4\) the scheme \(\text{Hilb}^{nm+1}({\mathbb P}^3)\) has more than two components, which are hard to describe. The component containing the rational normal curves contains a large subvariety where the tangent space is ``too big'', namely the points corresponding to an ACM \(n\)-fold structure on a line. In order to show that the component is singular at these points, it suffices to show that the set of ACM curves is irreducible. The paper under review begins with some general observations on ACM curves in \({\mathbb P}^n\) with Hilbert polynomial \(nm+1\), and then specializes to the case \(n=4\). The authors describe all possible (especially non-reduced) structures of ACM curves with Hilbert polynomial \(4m+1\), and they use this classification to show that any ACM curve is the limit of a family of rational normal curves, hence giving the desired irreducibility. They also show that the singular locus of this component has dimension greater than 6.