Line, Spiral, Dense in Space

Atakora Agoro, Alastair N. Fletcher · 2026

Read Line, Spiral, Dense in Space by Atakora Agoro, Alastair N. Fletcher (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-2605-28470.txt.

Introduction

Atakora Agoro and Alastair Fletcher prove a three-dimensional analogue of a density result for iterated exponential maps. Dobbs had shown that the second iterate of nearly every line in the complex plane beneath the exponential function fills the plane densely. Agoro and Fletcher show the same kind of density for the Zorich map's second iterate in three-dimensional space. Their short paper thus carries a planar iteration result into a spatial setting, with density as the shared conclusion.

Catalog record

CREATORAtakora Agoro, Alastair N. Fletcher
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE35687
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2605.28470

Browse the full library · About the Tower of Babel

Full text

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

Line, Spiral, Dense in Space Atakora Agoro, Alastair N. Fletcher Abstract . Dobbs proved that the second iterate of almost every line in the complex plane under the exponential function is dense in the plane. In this paper, we prove an analogous result for the second iterate of the Zorich map in ℝ 3 \mathbb{R}^{3} . 1. Introduction It is a standard homework exercise in complex analysis to show that the image of an oblique line under \exp is a spiral. It is much less standard to show that the image of almost all oblique lines under \exp\circ\exp is a dense curve in \mathbb{C} . This remarkable result is due to Dobbs [ 2 , Theorem 1] . Dobbs’ method was generalized by Vaskouski, Prochorov, and Sheshko [ 6 ] to cover a larger class of periodic functions in \mathbb{C} . The aim of this paper is to prove an analogous result for a quasiregular version of the exponential function in \mathbb{R}^{3} . Such maps are called Zorich maps, as they were first constructed by Zorich [ 7 ] and the dynamics of these maps have been well studied in the literature, see for example the work of Bergweiler [ 1 ] and Tsantaris [ 5 ] . We will fix a particular Zorich map \mathcal{Z}:\mathbb{R}^{3}\to\mathbb{R}^{3}\setminus\{0\} to work with in Section 2 . Theorem 1.1 . For almost every line L in \mathbb{R}^{3} , the curve (\mathcal{Z}\circ\mathcal{Z})(L) is dense in \mathbb{R}^{3} . The overarching strategy for proving Theorem 1.1 is the same as Dobbs, but there are more technical difficulties to overcome in \mathbb{R}^{3} . Given a ball U , we will study its second pre-image (\mathcal{Z}\circ\mathcal{Z})^{-1}(U) . Then instead of looking at a line to see whether it intersects this set, we will fatten it slightly, in a suitable sense, to give a cone and show that a certain proportion of this cone must intersect the second pre-image of the ball. We will then use measure theoretical considerations to obtain the result. For the exponential map, the real axis is forward invariant, which means there are certainly some lines which must be excluded. For our Zorich map \mathcal{Z} , there are analogous lines that will need to be excluded, so we certainly cannot obtain Theorem 1.1 for every line in \mathbb{R}^{3} . We leave open several related questions. First, Dobbs [ 2 , Theorem 3] showed that the image of every oblique line in \mathbb{C} was dense in \mathbb{C} under the third iterate of \exp . It seems plausible that the same might hold for \mathcal{Z} in \mathbb{R}^{3} . Moreover, there is a whole family of Zorich maps that one could study in the context of generalizing Theorem 1.1 and, more generally, the same result could plausibly hold for any strongly automorphic map in \mathbb{R}^{3} or higher dimensions. We, however, will satisfy ourselves with just proving Theorem 1.1 here. The structure of the paper is as follows. In Section 2 we will define the particular Zorich map we will use and recall various properties of it. In Section 3 , we will prove Theorem 1.1 . This paper arises from the Ph.D. dissertation of the first named author. The external examiner for the dissertation was Neil Dobbs, and both authors would like to thank him for many invaluable comments and suggestions. 2. Preliminaries 2.1. A Zorich map Let B denote the square (2.1) B=\left\{(x_{1},x_{2})\in\mathbb{R}^{2}:|x_{1}|,|x_{2}|\leq\frac{\pi}{2}\right\}, and let \mathbb{S}^{2}_{+} denote the upper hemisphere in \mathbb{R}^{3} given by \mathbb{S}^{2}_{+}=\left\{(x_{1},x_{2},x_{3})\in\mathbb{R}^{3}:x_{1}^{2}+x_{2}^{2}+x_{3}^{2}=1,x_{3}\geq 0\right\}. To build the Zorich map we will use, we start by defining the bi-Lipschitz map h:B\to\mathbb{S}^{2}_{+} given by (2.2) h(x_{1},x_{2})=\left(\frac{x_{1}\sin M(x_{1},x_{2})}{\sqrt{x_{1}^{2}+x_{2}^{2}}},\frac{x_{2}\sin M(x_{1},x_{2})}{\sqrt{x_{1}^{2}+x_{2}^{2}}},\cos M(x_{1},x_{2})\right), where M(x_{1},x_{2})=\max\{|x_{1}|,|x_{2}|\} . By reflecting in the sides of squares in the domain and in the equator of \mathbb{S}^{2} in the range, we obtain a doubly periodic map that, by an abuse of notation, we still call h from \mathbb{R}^{2} onto \mathbb{S}^{2} . The Zorich map \mathcal{Z} is then defined via (2.3) \mathcal{Z}(x_{1},x_{2},x_{3})=e^{x_{3}}h(x_{1},x_{2}). If we let B_{0} be the closed beam in \mathbb{R}^{3} defined by B_{0}=B\times\mathbb{R} , then it is clear that \mathcal{Z} maps B_{0} onto the closed upper half-space in \mathbb{R}^{3} with the origin removed. Moreover, if we set B_{1}=\left[\frac{\pi}{2},\frac{3\pi}{2}\right]\times\left[-\frac{\pi}{2},\frac{\pi}{2}\right]\times\mathbb{R} to be one of the beams neighbouring B_{0} , then the image of B_{1} under \mathcal{Z} is the closed lower half-space in \mathbb{R}^{3} with the origin removed. This produces a quasiregular map \mathcal{Z}:\mathbb{R}^{3}\to\mathbb{R}^{3}\setminus\{0\} . At this stage, we point out that we do not really need that \mathcal{Z} is quasiregular. It is only important for us that \mathcal{Z} is an analogue of \exp in \mathbb{R}^{3} . In fact, restricting to either the \{x_{1}=0\} plane of the \{x_{2}=0\} plane yields embedded copies of \exp . We refer to, for example, [ 4 , Appendix A2] for a proof of the fact that h defined in ( 2.2 ) is bi-Lipschitz and \mathcal{Z} is quasiregular. The branch set \mathcal{B}(\mathcal{Z}) is the set of points where \mathcal{Z} fails to be a local homeomorphism. From the construction, we see that the branch set consists of the lines at the edges of faces of the beams, that is \mathcal{B}(\mathcal{Z})=\bigcup_{j,k\in\mathbb{Z}}\left(\left\{\frac{\pi}{2}+j\pi\right\}\times\left\{\frac{\pi}{2}+k\pi\right\}\times\mathbb{R}\right). 2.2. The group of automorphisms It turns out that Zorich maps are strongly automorphic with respect to a discrete group of isometries. This means that there is such a group G with (a) \mathcal{Z}(g(x))=\mathcal{Z}(x) for any x\in\mathbb{R}^{3} and any g\in G , and (b) if \mathcal{Z}(x)=\mathcal{Z}(y) , then there exists g\in G with y=g(x) . It is clear from the construction that \mathcal{Z} is periodic with periods (2\pi,0,0) and (0,2\pi,0) . However, G also contains rotations by \pi about lines in the branch set. If we set g_{1}(x)=x+(2\pi,0,0) , g_{2}(x)=x+(0,2\pi,0) and g_{3}(x_{1},x_{2},x_{3})=(\pi-x_{1},\pi-x_{2},x_{3}), then G is generated by g_{1},g_{2} , and g_{3} . As \mathcal{Z} maps \operatorname{int}(B_{0}) bijectively onto the open upper half-space, and maps \operatorname{int}(B_{1}) bijectively onto the open lower half-space, it follows that \operatorname{int}(B_{0}\cup B_{1}) together with some of its boundary forms a fundamental domain for the action of G that we call \Omega . We will see later that this domino behaviour where a fundamental domain is naturally split into two regions has large consequences for our analysis. 2.3. Mapping properties of our Zorich map Here we record some important mapping properties of \mathcal{Z} that we will need later. Lemma 2.1 . For t\in\mathbb{R} , let H_{t} be the plane H_{t}=\{(x_{1},x_{2},t)\in\mathbb{R}^{3}:x_{1},x_{2}\in\mathbb{R}\}. Then \mathcal{Z} maps H_{t} onto the sphere S(0,e^{t})=\{x\in\mathbb{R}^{3}:|x|=e^{t}\} . Conversely, \mathcal{Z}^{-1}(S(0,r))=H_{\ln r} for any r>0 . This follows directly from the construction of \mathcal{Z} and so we omit the proof. Lemma 2.2 . Let t>0 and let S_{t}\subset B_{0} be defined via S_{t}=\{(x_{1},x_{2},x_{3})\in B_{0}:e^{x_{3}}\cos M(x_{1},x_{2})=t\}. Then \mathcal{Z} maps S_{t} bijectively onto the plane H_{t} . Proof. As \mathcal{Z} is injective on B_{0} , it is enough to find the pre-image of H_{t} under \mathcal{Z} contained in B_{0} . By ( 2.2 ) and ( 2.3 ), this follows by solving e^{x_{3}}\cos M(x_{1},x_{2})=t . Re-writing this expression, we also obtain that S_{t}=\left\{(x_{1},x_{2},x_{3})\in B_{0}:x_{3}=\ln\left(\frac{t}{\cos M(x_{1},x_{2})}\right)\right\}. ∎ For t<0 , we define S_{t}\subset B_{1} to be the reflection of S_{-t} across the common face of B_{0} and B_{1} . Lemma 2.3 . Let t\in\mathbb{R} . Then: (i) for t>0 , \mathcal{Z}^{-1}(H_{t}) is \bigcup_{g\in G}g(S_{t}) , (ii) for t<0 , \mathcal{Z}^{-1}(H_{t}) is \bigcup_{g\in G}g(S_{t}) ,

Read the complete text