On a characterization of path connected topological fields (Q2318355)

Let us first recall some notations and terminologies used throughout this report. It is well known that a ring \(R\) is said to be a topological ring if it is a topological space and the mappings \((x,y)\mapsto x-y\) and \((x,y)\mapsto xy\) are continuous. A topological field \(F\) is a topological ring where, in addition, it is required that the inverse mapping \(a\mapsto a^{-1}\) is continuous on \(F\setminus \{0\}\). The ring of all continuous functions from a topological space \(X\) to a topological field \(F\) is denoted by \(C(X,F)\); \(C(X,\mathbb{R})\) is denoted by \(C(X)\). Basic classical results concerning rings of continiuous real-valued functions can be found in [\textit{L. Gillman} and \textit{M. Jerison}, Rings of continuous functions. Princeton-Toronto-London-New York: D. Van Nostrand Company, Inc. (1960; Zbl 0093.30001)], and a comprehensive survey of the results in the context of rings of continuous functions with values in a topological ring other than the real field can be found in [\textit{E. M. Vechtomov}, J. Math. Sci., New York 78, No. 6, 702--753 (1996; Zbl 0868.46018)]. A topological space \(X\) is said to be path connected if for every two points \(x,y\in X\), there is a continuous path \(\lambda: [0,1]\rightarrow X\) joining \(x\) and \(y\), and to be arcwise connected if \(\lambda\) may be always chosen as a homeomorphism between \([0,1]\) and its image. It is well known that, in the context of topological fields, path connectedness coincides with arcwise connectedness. A topological space \(X\) is said to be \(F\)-regular if the collection \(\{Z(f): f\in C(X,F)\}\) separates points from closed sets of \(X\) in which \(Z(f)=\{x\in X: f(x)=0\}\). Note that zero-dimensional spaces (a Hausdorff space containing a base of clopen sets is called zero-dimensional) are exactly \(F\)-regular spaces for an arbitrary disconnected field \(F\), and Tychonoff spaces are exactly \(F\)-regular spaces for arcwise connected fields, see [\textit{E. M. Vechtomov}, Math. Notes 55, No. 6, 568--579 (1994; Zbl 0841.54013); translation from Mat. Zametki 55, No. 6, 32--49 (1994)]. The classical theorem of Gelfand and Kolmogoroff states that \(\mathrm{Max}(C(X))\), the space of maximal ideals of \(C(X)\) equipped with the hull-kernel topology, is homeomorphic to the Stone-Čech compactification of \(X\), \(\beta X\), under a homeomorphism whose restriction to the subspace consisting of fixed maximal ideals is a homeomorphism onto \(X\). Note that fixed maximal ideals of \(C(X)\) are precisely the ideals \(M_x=\{f\in C(X): f(x)=0\}\) for \(x\in X\), and it is well known that a Tychonoff space \(X\) is compact if and only if every maximal ideal of \(C(X)\) is fixed. The Gelfand and Kolmogoroff theorem has been generalized to the rings \(C(X,F)\) by \textit{E. M. Vechtomov} in [loc. cit.] and [Russ. Math. Surv. 47, No. 5, 171--172 (1992; Zbl 0795.54027); translation from Usp. Mat. Nauk 47, No. 5, 171--172 (1992)]. In the first of these papers, it has been stated (Theorem 5) that for any disconnected (resp., arcwise connected) topological field \(F\) and any zero-dimensional (resp., Tychonoff) space \(X\) the ring \(C(X,F)\) satisfies the Gelfand-Kolmogoroff theorem with respect to the Banaschewski compactification \(\beta_0X\) (resp., the Stone-Čech compactification). Note that a ring \(C(X,F)\) is said to satisfy the Gel{f}and-Kolmogoroff theorem, briefly, GKT, with respect to the compactification \(K\) of \(X\) if the mapping \(p\mapsto M^p\) is a homeomorphism between the spaces \(K\) and \(\mathrm{Max}(C(X,F))\) in which \(M^p\) denotes \(\{f\in C(X,F): p\in \mathrm{cl}_{K}Z(f)\}\). This mapping is called the Gelfand mapping and, following the notations of the paper under review, we denote it by \(I_F\). In the paper under review, using the Gelfand mapping \(I_F\), for the case that \(X\) is a compact Hausdorff space, a characterization of path-connected topological fields is investigated. More precisely, it is proved that, for every compact Hausdorff topological space \(X\), the Gelfand mapping \(I_F\) is a homeomorphism onto \(X\) if and only if \(F\) is path connected, Theorem 4.1 of the paper. We next show that this fact easily follows from the results of \textit{E. M. Vechtomov} [loc. cit.], and moreover, could be generalized for non-compact topological spaces. However, the paper investigates a new proof for the fact. For the sake of the reader, we cite some basic statements of [\textit{E. M. Vechtomov}, loc. cit.] which are used for proving the mentioned fact. Note that a topological field \(F\) is said to satisfy the GKT if for each \(F\)-regular topological space \(X\), the ring \(C(X,F)\) satisfies the GKT with respect to \(\beta_FX\), see [\textit{G. Bachman} et al., Trans. Am. Math. Soc. 204, 91--112 (1975; Zbl 0299.54016)] for details about the compactification \(\beta_FX\). Theorem 1. A ring \(C(X,F)\) satisfies the GKT with respect to some compactification (Hausdorff compactification) if and only if \(X\) is \(F\)-regular and each \(M^p\) is a \(z\)-ideal. Theorem 2. Every disconnected or linearly connected topological field satisfies the Gelfand-Kolmogorov theorem. The next statement could be considered as a generalization of Theorem 4.1 of the paper under review. Theorem 3. Let \(F\) be topological field. Then the Gelfand map \(I_F\) is a homeomorphism for every Tychonoff topological space \(X\) if and only if \(F\) is path connected. (\(\Rightarrow\)) Let \(X\) be a non zero-dimensional Hausdorff space. By the hypothesis, the Gelfand mapping \(I_F\) is a homeomorphism and thus, by Theorem 1, \(X\) is \(F\)-regular. Therefore, as \(X\) is not zero-dimensional, \(F\) should be path-connected. (\(\Leftarrow\)) Let \(F\) be a path-connected topological field and \(X\) be a Tychonoff space. As \(X\) is a Tychonoff space, it easily follows that \(\beta_FX=\beta X=X\) and thus, by Theorem 2, the Gelfand mapping \(I_F\) would be a homeomorphism.