Minimal algebraic solutions of the sixth equation of Painlevé
Robert Conte 1. Université Paris-Saclay, ENS Paris-Saclay, CNRS,Centre Borelli, F-91190 Gif-sur-Yvette, France.2. Department of Mathematics, The University of Hong Kong,Pokfulam, Hong Kong · 2025
Read Minimal algebraic solutions of the sixth equation of Painlevé by Robert Conte 1. Université Paris-Saclay, ENS Paris-Saclay, CNRS,Centre Borelli, F-91190 Gif-sur-Yvette, France.2. Department of Mathematics, The University of Hong Kong,Pokfulam, Hong Kong (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2504-08287.txt.
Introduction
Robert Conte studies the forty-eight exceptional algebraic solutions of the sixth Painlevé equation, a famous nonlinear equation in mathematical physics. For each solution, he builds the algebraic curve that the solution traces, aiming at a curve whose degree is as low as possible. He then gives a best-possible parametric form for each curve. The degree matches the number of branches of the solution in all but fifteen cases.
Catalog record
| CREATOR | Robert Conte 1. Université Paris-Saclay, ENS Paris-Saclay, CNRS,Centre Borelli, F-91190 Gif-sur-Yvette, France.2. Department of Mathematics, The University of Hong Kong,Pokfulam, Hong Kong |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 44 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2504-08287
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Minimal algebraic solutions of the sixth equation of Painlevé Robert Conte 1. Université Paris-Saclay, ENS Paris-Saclay, CNRS,Centre Borelli, F-91190 Gif-sur-Yvette, France.2. Department of Mathematics, The University of Hong Kong,Pokfulam, Hong Kong Abstract For each of the forty-eight exceptional algebraic solutions u ( x ) u(x) of the sixth equation of Painlevé, we build the algebraic curve P ( u , x ) = 0 P(u,x)=0 of a degree conjectured to be minimal, then we give an optimal parametric representation of it. This degree is equal to the number of branches, except for fifteen solutions. Keywords : sixth Painlevé equation, algebraic solutions. PACS 02.30.Hq AMS MSC 2000 33E17 1 Introduction and motivation Apart the transcendental general solution which characterizes it, the sixth equation {\rm P_{\rm VI}} of Painlevé \displaystyle{\hskip-56.9055pt}\frac{\hbox{d}^{2}u}{\hbox{d}x^{2}}=\frac{1}{2}\left[\frac{1}{u}+\frac{1}{u-1}+\frac{1}{u-x}\right]\left(\frac{\hbox{d}u}{\hbox{d}x}\right)^{2}-\left[\frac{1}{x}+\frac{1}{x-1}+\frac{1}{u-x}\right]\frac{\hbox{d}u}{\hbox{d}x} \displaystyle{\hskip-56.9055pt}\phantom{123456}+\frac{u(u-1)(u-x)}{2x^{2}(x-1)^{2}}\left[\theta_{\infty}^{2}-\theta_{0}^{2}\frac{x}{u^{2}}+\theta_{1}^{2}\frac{x-1}{(u-1)^{2}}+(1-\theta_{x}^{2})\frac{x(x-1)}{(u-x)^{2}}\right], \displaystyle{\hskip-56.9055pt}(2\alpha,-2\beta,2\gamma,1-2\delta)=(\theta_{\infty}^{2},\theta_{0}^{2},\theta_{1}^{2},\theta_{x}^{2}), (1) admits exactly 48 exceptional algebraic solutions u(x) , in the sense that they are neither those mentioned by Picard [ 1 , p. 298] when the four constant parameters \theta_{j} vanish, nor some particular solutions of the Riccati equation admitting {\rm P_{\rm VI}} as a differential consequence for some constraint between the \theta_{j} ’s. They were discovered by various authors : Hitchin [ 2 ] , Dubrovin [ 3 ] , Dubrovin and Mazzocco [ 4 ] , Kitaev [ 5 ] [ 6 ] , Andreev and Kitaev [ 7 ] , and all the others by Boalch [ 8 ] [ 9 ] [ 10 ] [ 11 ] . The proof that no other such solutions exist is due to Lisovyy and Tykhyy [ 12 ] . These 48 solutions are in fact equivalence classes of a group which leaves {\rm P_{\rm VI}} form-invariant without constraining the \theta_{j} ’s, group made of: the 4 sign changes of the \theta_{j} ’s, the 4! homographies of (u,x) and a unique birational transformation (see Appendix). Let us denote P(u,x)=0 the algebraic curve representing an element of one of these equivalence classes. Its genus g and number of branches b (degree of the polynomial P in u ) are invariant under the group, but its degree d (global degree of P in u and x ) is not. In order to simplify the writing of the solutions, it is therefore convenient to lower this degree d , ideally to its minimal value b , by the repeated action of elements of the group. Remark . The above notation \mbox{\boldmath{$\theta$}}=(\theta_{\infty},\theta_{0},\theta_{1},\theta_{x}) for the sequence of the four monodromy exponents is the natural choice for at least three reasons. The first one is the writing of {\rm P_{\rm VI}} in elliptic coordinates [ 13 , 14 , 15 ] \displaystyle\frac{\hbox{d}^{2}U}{\hbox{d}X^{2}}=\frac{(2\omega)^{3}}{\pi^{2}}\sum_{j=\infty,0,1,x}\theta_{j}^{2}\wp^{\prime}(2\omega U+\omega_{j},g_{2},g_{3}), (2) obviously form-invariant under any permutation of the four half-periods \omega_{j} of the elliptic function \wp ; defining \theta_{\infty} with a shift of unity by 2\alpha=(\theta_{\infty}-1)^{2} would break this invariance. The second one is the classical confluence [ 14 ] between the {\rm P_{\rm n}} ’s, in which the four singularities (\infty,0,1,x) of u become successively (\infty,0,1) , (\infty,0) , (\infty) , which defines an order in the sequence (\theta_{\infty},\theta_{0},\theta_{1},\theta_{x}) ; reordering this sequence into (\theta_{0},\theta_{x},\theta_{1},\theta_{\infty}-1) [ 8 , Eq. (4)] or into (\theta_{0},\theta_{1},\theta_{x},\theta_{\infty}-1) [ 12 , p. 125] would make one forget about this natural ordering. The third one is the choice of Painlevé [ 14 ] for the Greek alphabetical order \alpha,\beta,\gamma,\delta . 2 Statement of the problem The goal of the present paper is thefore to solve the following problem. Problem . For each of the 48 algebraic solutions, find a representative in its equivalence class whose degree d equals the number of branches b . As an example, consider one of the two solutions with b=5 branches, the one found by Kitaev [ 5 , Eq. (3.3)] which bears the number I21 in the tables de Boalch [ 8 ] and 2 in those of Lisovyy and Tykhyy [ 12 ] . Its genus is g=0 , its degree the nonoptimal value d=b+1 and its number of terms is 12. An homography followed by the birational transformation simplifies it to b=d=5 containing only 4 terms, \displaystyle\left\{\begin{array}[]{ll}\displaystyle{\mbox{\boldmath{$\theta$}}=\frac{(3,1,2,1)}{5},P=9(2-x)u^{5}+15(x-4)u^{4}+5(10-x^{2}+9x)u^{3}-15(5x+3x^{2})u^{2}}\\ \displaystyle{\phantom{12345678901234567890}+45x^{2}u-x^{2}-8x^{3},}\\ \displaystyle{\mbox{\boldmath{$\theta$}}=\frac{(1,1,2,3)}{5},P=(9x-8)u^{5}-5u^{4}x-5x(x-3)u^{3}-5x(3x-1)u^{2}+5x^{2}u-9x^{2}+8x^{3},}\\ \displaystyle{\mbox{\boldmath{$\theta$}}=\frac{(0,0,1,2)}{5},P=u^{5}-5xu^{3}+5xu^{2}-x^{2}.}\end{array}\right. The sought simplification therefore consists in removing any fixed pole of u (in the above example I21, the fixed pole x=2 ), so that u admits as only singularities the three critical fixed points x=\infty,0,1 . For such removable fixed singularities, see for example the figures in [ 16 ] . There exists another transformation [ 17 ] , called folding transformation, and it is unique [ 18 ] , which leaves {\rm P_{\rm VI}} form-invariant, but this is at the expense of two constraints between the \theta_{j} ’s. A solution \mbox{\boldmath{$\theta$}}=(0,0,2a,2b) ( modulo permutations and sign changes) is folded into another solution \mbox{\boldmath{$\theta$}}=(a,a,b,b) , an operation symbolized as “unfolded \leq folded” par Boalch. Let us recall the partition of the 48 algebraic solutions in three disjoint types [ 19 , p. 233] : 1. (genus zero) three solutions which depend on at most two arbitrary \theta_{j} ’s; 2. (genus zero or one) thirty solutions having rational, non-arbitrary \theta_{j} ’s, inequivalent under the folding transformation [ 17 ] ; 3. (genera 0, 1, 2, 3, 7) fifteen transformed of seven out of the thirty ones by folding(s). The final result is the following. Conjecture . Only 7 of the 30 unfolded solutions, and 8 of the 15 folded ones display a nonzero difference d-b , and this difference is minimal. Before the present work, only 4 of the 30 unfolded solutions, and 3 of the 15 folded ones displayed a minimal d-b . The main reason is precisely the elegance of the method of Boalch. That method makes him find rational numbers \theta_{j} whose denominators have a minimal gcd [ 8 , Table 2] , and it happens that the birational transformation doubles this gcd in most cases but that, surprisingly, it greatly simplifies (in the sense above defined) the representative. Table 1 displays numerous such examples : I22, I23, O09, I24, I25, etc. Several facts support this conjecture. 1. Existence of several sets of siblings (in the sense defined by Boalch [ 8 ] , i.e. whose all members have by definition the same values of b and x ), one member having an optimal degree d=b and the others degrees d\geq b+1 impossible to lower by any transformation. 2. Probable nonexistence, for the solution I50 with b=40 branches, of a representative d-b\leq 5 whose all \theta_{j} ’s would be equal, so as to preserve its property to be doubly folded. Indeed, the representatives whose common value of the four \theta_{j} ’s is respectively 3/20,7/20,13/20,17/20 have degrees 46,54,78,102 and a number of terms equal to 335,663,1647,2631 . 3. Failure to find a representative d=b of the solution with the smallest b having a degree d\not=b (solution labeled K= Klein), after a triple loop on the 2^{4} sign changes, the 4! homographies and five birational transformations (three months of computation). 3 Results They take two forms. 1. The solution of the above stated problem, summarized in Table 1 : among the 30+15 solutions, the 7+8 without a representative d=b have a minimal number d-b of fixed poles equal to six for I50, I52, two for I38, 238,