On the classification of \((M-2)\)-curves of order 8 (Q3978428)

It is well known that the problem of classifying the position in \(\mathbb{R} \mathbb{P}^ 2\) of branches of nonsingular curves of order \(m\) when \(m \leq 5\) is easily solved by applying Bézout's theorem and Harnack's constructions, and when \(m=6\) it was posed by Hilbert in his 16th problem and solved in 1969 by Gudkov. For \(m=7\) the problem was solved by Viro. There he made important progress in the question of \(M\)-curves of order 8. In this paper we study the position in \(\mathbb{R}\mathbb{P}^ 2\) of branches of a curve of order 8 consisting of 20 ovals. Such positions are called real \((M-2)\)-schemes of order 8, and are denoted by \(S_ 8^{M-2}\). The topological classification of schemes \(S_ 8^{M-2}\) has particular interest, since it is known that when \(m<8\) all logically possible (that is, topologically not contradicting Bézout's theorem in their intersection with curves of lower order) real schemes \(S_ m^{M-2}\) can be realized. When \(m=8\) this is not so: as Viro observed from Kharlamov's congruence [cf. \textit{V. A. Rokhlin}, Russ. Math. Surv. 33, 85-98 (1978); translation from Usp. Mat. Nauk 33, No. 5 (203), 77-89 (1978; Zbl 0437.14013); subsection 3, 4] and \textit{T. Fiedler}'s theorem on complex orientations [Math. USSR, Izv. 21, 161-170 (1983); translation from Izv. Akad. Nauk SSSR, Ser. Mat. 46, 853-863 (1982; Zbl 0522.14014); theorem 3] it follows that 9 of the schemes \(S_ 8^{M-2}\) cannot be realized. By direct calculation it is not difficult to see that, of all the 419 logically possible pairwise distinct schemes \(S_ 8^{M-2}\), 100 satisfy Kharlamov's congruence, but 10 of these 100 schemes cannot be realized. Of the remaining 409 schemes I have succeeded in realizing 138, of which 29 satisfy Kharlamov's congruence.