Trees and generalised inverse limits on intervals (Q1800233)

A generalized inverse limit (GIL) is defined for sequences \(\{(X_{n},f_{n})\}_{n=1}^{\infty}\), where each \(X_{n}\) is a compact metric space and \(f_{n}\) is an upper semi-continuous valued function \(f_{n}:X_{n+1}\rightarrow 2^{X_{n}}\), in the following way: \[ \lim_{\leftarrow}(X_{n},f_{n})=\{\{x_{n}\}_{n=1}^{\infty}\in \prod_{n=1}^{\infty}X_{n}:x_{n}\in f_{n}(x_{n+1})\text{ for all }n\in \mathbb{N}\}. \] The simplest GILs are those in which each \(X_{n}\) is the unit interval \([0,1] \), each \(f_{n}\) is represented by a closed subset \(M_{n}\) in the unit square \([0,1]^{2}\) and each \(M_{n}\) projects onto each of its two coordinates. In this case, the inverse limit can be written as \[ \lim_{\leftarrow}([0,1],M_{n})=\{\{x_{n}\}_{n=1}^{\infty}\in [ 0,1]^{\infty}:(x_{n+1},x_{n})\in M_{n}\text{ for all }n\in \mathbb{N}\}. \] In the case that each \(M_{n}\) is homeomorphic to \([0,1]\) (\(M_{n}\) is an arc), the sequence \(\{([0,1],M_{n})\}_{n=1}^{\infty}\) is called a full-arc sequence. A continuum is a compact connected metric space. A very interesting problem in the area of GILs is: for which continua \(X\) does there exist a sequence \(\{M_{n}\}_{n=1}^{\infty}\) such that \(X\) is homeomorphic to \(\lim_{\leftarrow}([0,1],M_{n})\)? In fact, for some time, it was an open question whether every continuum can be given as such a GIL. Only recently, \textit{S. Greenwood and R. Suabedissen} [Discrete Contin. Dyn. Syst. 37, No. 11, 5693--5706 (2017; Zbl 1387.54017)] solved this question by proving that the only compact \(2\)-manifold \(Z\) for which there exists a sequence \(\{M_{n}\}_{n=1}^{\infty}\) such that \(Z\) is homeomorphic to \(\lim_{\leftarrow}([0,1],M_{n})\) is the torus. In the paper under review, the authors study the problem of determining which continua are GILs of full-arc sequences. Using clever constructions they determine that there exist two specific arcs \(K\) and \(L\) in \([0,1]^{2}\) such that every finite tree \(T\) is homeomorphic to a limit of the form \(\lim_{\leftarrow}([0,1],M_{n})\), where \(M_{n}\in \{K,L\}\) for all \(n\in \mathbb{N}\). They also study how to obtain some infinite trees as GILs of full-arc sequences. The paper ends with some interesting questions, we mention two of them. Question 3. Can every tree-like continuum be obtained as a GIL of a full-arc sequence? Question 5. Is there a finite set \(S\) of full-arc functions such that if \(C\) is a tree-like continuum that can be obtained as a GIL for a full-arc sequence, then \(C\) can be obtained as a GIL with bonding function members of \(S\)?