Any six points on the Riemann sphere can be split into three pairs by a triple of disjoint discs

Matvey Smirnov · 2026

Read Any six points on the Riemann sphere can be split into three pairs by a triple of disjoint discs by Matvey Smirnov (2026) 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-2604-00351.txt.

Introduction

Smirnov proves that any six points on the Riemann sphere can be divided into three pairs, each pair enclosed by its own disc, with the three discs disjoint. The result applies whether the discs are closed or open. It also validates a recent numerical method for evaluating genus-2 Kleinian hyperelliptic functions, showing the method works for every complex curve of that genus.

Catalog record

CREATORMatvey Smirnov
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE92259
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2604.00351

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Any six points on the Riemann sphere can be split into three pairs by a triple of disjoint discs Matvey Smirnov Abstract . We prove that for any six points on the Riemann sphere there exist three disjoint closed (or open) discs, each of which contains exactly two of the six distinguished points. This statement shows that recently proposed method to numerically evaluate Kleinian hyperelliptic functions of genus 2 is applicable to any complex curve of genus 2. Keywords. Elementary geometry, analytic geometry, Riemann sphere. 1991 Mathematics Subject Classification 51M04, 51N20 1. Introduction In this paper we denote the Riemann sphere by \CC\PP(1) . As usual we identify the complex plane \CC with the affine part of \CC\PP(1) , so \CC\PP(1)=\CC\cup\{\infty\} . Also let \mathbb{D}=\{z\in\CC:|z|<1\} denote the open unit disc, and \mathbb{T}=\{z\in\CC:|z|=1\}=\partial\mathbb{D} denote the unit circle. Recall that biholomorphic mappings S:\CC\PP(1)\to\CC\PP(1) are precisely the Möbius transformations, i.e. S(z)=\dfrac{az+b}{cz+d}\mathchar 59\relax where ad-bc\neq 0 . For the basic properties of Riemann sphere and Möbius transformations we refer to [ 2 , Chap. VI] . By a closed (resp. open ) disc in \CC\PP(1) we call a set D\subset\CC\PP(1) that is the image of \overline{\mathbb{D}} (resp. \mathbb{D} ) with respect to a Möbius transformation. That is, a closed disc in \CC\PP(1) is either a closed disc in \CC (with arbitrary center and radius), or the complement of an open disc in \CC , or a closed half-plane (with \infty ). The open discs are described similarly. The aim of this paper is to prove the following fact. Theorem 1.1 . Let A\subset\CC\PP(1) be any six-element set. Then there exist three disjoint closed discs D_{1}\mathord{\mathchar 59\relax}D_{2}\mathord{\mathchar 59\relax}D_{3}\subset\CC\PP(1) such that D_{j}\cap A consists of exactly two points for all j\in\{1\mathord{\mathchar 59\relax}2\mathord{\mathchar 59\relax}3\} . It is easy to see that the following statements are equivalent to Theorem 1.1 . Theorem 1.2 . Let A\subset\CC\PP(1) be any six-element set. Then there exist three disjoint open discs D_{1}\mathord{\mathchar 59\relax}D_{2}\mathord{\mathchar 59\relax}D_{3}\subset\CC\PP(1) such that D_{j}\cap A consists of exactly two points for all j\in\{1\mathord{\mathchar 59\relax}2\mathord{\mathchar 59\relax}3\} . Theorem 1.3 . Let S\subset\RR^{3} be the Euclidean 2-dimensional sphere, i.e. S=\{(x\mathord{\mathchar 59\relax}y\mathord{\mathchar 59\relax}z)\in\RR^{3}:x^{2}+y^{2}+z^{2}=1\} , and let A\subset S be any six-element set. Then there exist three disjoint open (or closed) discs B_{1}\mathord{\mathchar 59\relax}B_{2}\mathord{\mathchar 59\relax}B_{3}\subset S (here by discs we mean balls in metric space S equipped with Euclidean distance) such that B_{j}\cap A consists of exactly two points for all j\in\{1\mathord{\mathchar 59\relax}2\mathord{\mathchar 59\relax}3\} . The proof of equivalence of Theorems 1.1 - 1.3 is obvious and we omit it. The motivation for this problem comes from the paper [ 5 ] , where such splittings of six points into three pairs were used to formulate the algorithm that computes Kleinian hyperelliptic functions (for definitions see, e.g. [ 1 ] , [ 6 ] ) associated with complex curves of genus 2. This algorithm exploits the recursive reduction of the problem from a given curve to an isogenous curve, which can be found by a construction due to Richelot (see, e.g. [ 4 ] , [ 7 , Chap. 8] ). The construction of an isogenous curve is not unique: there are (in the generic case) 15 options that correspond to partitions of the set of Weierstrass points of the initial curve into three pairs. In [ 5 ] it was shown that given three disjoint open discs such that each of them contains exactly two of the Weierstrass points, then the Weierstrass points of corresponding Richelot isogenous curve are again split into three pairs by the same triple of discs. This fact allows one to iterate the Richelot’s construction and obtain effective numerical procedures for periods of canonical differentials on a curve of genus 2 and for associated Kleinian functions. However, in [ 5 ] it was not proved that such triple of discs always exists. The rest of the paper is devoted to the proof of Theorem 1.1 . The structure of this proof (and the sections of this paper) is discussed in the next section. Here we only note that the proof is very elementary and based on it one can quite easy formulate an algorithm that constructs a suitable triple of discs for any six-point set E\subset\CC\PP(1) . We shall not formulate such an algorithm. 2. Basic reduction of the problem and the idea of the proof For brevity we call a set A\subset\CC\PP(1) that contains exactly six points splittable if there exist three disjoint closed discs D_{1}\mathord{\mathchar 59\relax}D_{2}\mathord{\mathchar 59\relax}D_{3}\subset\CC\PP(1) such that A\cap D_{j} contains exactly two points for all j\in\{1\mathord{\mathchar 59\relax}2\mathord{\mathchar 59\relax}3\} . With this terminology we can formulate Theorem 1.1 as the statement that all six-element sets A\subset\CC\PP(1) are splittable. Lemma 2.1 . Let A\subset\CC\PP(1) be a six-element set and S:\CC\PP(1)\to\CC\PP(1) be a Möbius transformation. Then A is splittable if and only if S(A) is splittable. Proof. This is obvious, since Möbius transformations are bijective and map closed discs to closed discs. ∎ Using Lemma 2.1 we can without loss of generality impose additional conditions on a six-element set A in order to verify its splittability. To formulate such conditions we consider the set \Omega=\{z\in\CC:|z-1|<2\}\cup\{z\in\CC:|z+1|<2\}=(1+2\mathbb{D})\cup(-1+2\mathbb{D})\mathchar 59\relax which is shown on Fig. 1 . \sqrt{3}i -\sqrt{3}i 1 -1 Figure 1 . The set \Omega ; the boundary of \Omega is shown by the solid line. Definition 2.2 . We call a three-element set E\subset\CC distinguished if E\cap\Omega=\emptyset and |z-w|\geq 2 for all z\mathord{\mathchar 59\relax}w\in E , z\neq w . Proposition 2.3 . Assume that for any distinguished three-element set E\subset\CC the set E\cup\{1\mathchar 59\relax-1\mathchar 59\relax\infty\}\subset\CC\PP(1) is splittable. Then any six-element set A\subset\CC\PP(1) is splittable. Proof. Let A\subset\CC\PP(1) be arbitrary six-element set. At first we can find a Möbius transformation S such that \infty\in S(A) . After that we can choose distinct z\mathord{\mathchar 59\relax}w\in S(A)\setminus\{\infty\} with smallest possible distance and find an affine transformation T , such that T(z)=-1 and T(w)=1 . Thus, we arrive at the six-element set T(S(A)) such that \{-1\mathchar 59\relax 1\mathchar 59\relax\infty\}\subset T(S(A)) and 2 (which is the distance between -1 and 1 ) is the smallest possible distance between distinct elements of T(S(A))\setminus\{\infty\} . Now it is easy to verify that the set E=T(S(A))\setminus\{-1\mathchar 59\relax 1\mathchar 59\relax\infty\} is distinguished, so T(S(A)) is splittable by assumption (and so is A by Lemma 2.1 ). ∎ Now the problem is to show splittability only for the sets of the form A=E\cup\{-1\mathord{\mathchar 59\relax}1\mathord{\mathchar 59\relax}\infty\} , where E is distinguished. Since A constains \infty it will be customary to us to reformulate splittability in this case. From now on we shall denote the convex hull of a set X\subset\CC by \conv X . Lemma 2.4 . Let e_{1}\mathchar 59\relax\dots\mathchar 59\relax e_{5} be any five distinct points in \CC . Assume further that there exist closed bounded disjoint discs F_{1}\mathchar 59\relax F_{2}\subset\CC such that e_{1}\mathchar 59\relax e_{2}\in F_{1} , e_{3}\mathchar 59\relax e_{4}\in F_{2} and e_{5}\notin\conv(F_{1}\cup F_{2}) . Then the set \{e_{1}\mathchar 59\relax e_{2}\mathchar 59\relax e_{3}\mathchar 59\relax e_{4}\mathchar 59\relax e_{5}\mathchar 59\relax\infty\} is splittable. Proof. Since e_{5}\notin\conv(F_{1}\cup F_{2}) and the set \conv(F_{1}\cup F_{2}) is convex and compact, there exists a closed half-plane H\subset\CC\PP(1) such that e_{5}\in H and H\cap\conv(F_{1}\cup F_{2})=\emptyset (see [ 3 , Theorems 1.12 and 1.14] ). Clearly, F_{1} , F_{2} , and H constitute a triple of closed

Read the complete text