Recurrent networks and a theorem of Nash-Williams (Q756280)

\textit{C. St. J. A. Nash-Williams} [Proc. Camb. Philos. Soc. 55, 181-194 (1959; Zbl 0100.136)] showed, using random walk concepts and techniques, that an infinite, locally finite (electrical) network is recurrent if and only if the vertices can be partioned into finite sets \(P_ 0,P_ 1,..\). in such a manner that there are no edges between \(P_ i\) and \(P_ j\) unless \(| i-j| \leq 1\) and that \(\sum^{\infty}_{i=1}\sum^{\infty}_{j=1}1/c(P_ i,P_ j)=\infty,\) where \(c(P_ i,P_ j)\) is the sum of conductances of edges between \(P_ i\) and \(P_ j\). The author proves Nash-Williams' result without the assumption of local finiteness, so that vertices may have infinite degree. In addition, the result is shown to be related to recurrence and the Dirichlet principle for reversible Markov chains with countable space and to critical phenomena for certain infinite particle systems.