Completion of the variety of conics with respect to contact conditions. I (Q583318)

Denote by X the blow-up of the \({\mathbb{P}}^ 5\) of plane conics along the Veronese surface which parametrizes the double lines. One checks easily that X is the closure of the graph of the correspondence between each smooth conic and its dual. To find the number of conics tangent to five general conics [see for example \textit{S. L. Kleiman} in Studies in algebraic geometry, MAA Stud. Math. 20, 117-138, 142-143 (1980; Zbl 0444.14001)], we pass to X so as to be able to obtain a proper intersection for the five divisors of conics tangent to each of the given conics. We can do this by an elementary flatness argument even if there were infinitely many orbits [the reviewer in Algebraic Geometry, Proc. Conf., Sundance/Utah 1986, Lect. Notes Math. 1311, 235-252 (1988; Zbl 0683.14003)] provided that four of the divisors to be intersected meet each orbit properly. (For tangency, this evidently fails on \({\mathbb{P}}^ 5.)\) For more subtle conditions than tangency, properness may fail, typically when the varieties to be intersected have larger than expected intersections with the \(Halphen\quad locus\quad H,\) which parametrizes the degenerate complete conics consisting of double lines equipped with coincident vertices. To proceed further, we must blow up X along H, determine the resulting intersection ring, and investigate how the subvarieties representing the given conditions meet the orbits under the action of \(G=PGL(2)\), for which the given conics are to be made general. The present author seem to be unaware of the recent monograph by \textit{E. Casas-Alvéro} and \textit{S.-Xambó-Descamps} ``The enumerative theory of conics after Halphen'', Lect. Notes Math. 1196 (1986; Zbl 0626.14037) which treats these questions thoroughly; the paper under review covers part of the same ground, particularly the behavior of the various conditions with respect to the G-action), using classical-style arguments. The study of conics, an explicitly special case, illustrates (though the paper under review does not choose to do so) some features of the more difficult and still not fully understood geometry of complete quadrics in \({\mathbb{P}}^ n\). Two recent papers at this general level deserve attention: \textit{A. Thorup} and \textit{S. Kleiman} in Algebraic Geometry, Proc. Conf., Sundance/Utah 1986, Lect. Notes Math. 1311, 253-320 (1988; Zbl 0668.14030) and \textit{C. De Concini}, \textit{M. Goresky}, \textit{R. MacPherson} and \textit{C. Procesi}, ``On the geometry of quadrics and their degenerations'', Comment. Math. Helv. 63, No.3, 337-413 (1988)].