login/create account
    Infinite uniquely hamiltonian graphs
? (Originally appeared as [M].)
Let 
 be a locally finite infinite graph and let 
 be the set of ends of~
. The Freudenthal compactification of 
 is the topological space 
 which is obtained from the usual topological space  of the graph, when viewed as a 1-dimensional cell complex, by adding all points of 
 and setting, for each end 
, the basic set of neighborhoods of 
 to consist of sets of the form  
, where 
 ranges over the finite subsets of 
, 
 is the component of 
 containing all rays in 
, the set 
 contains all ends in 
 having rays in  
, and 
 is the union of half-edges 
, one for every edge 
 joining 
 and 
. We define a hamilton circle in 
 as a homeomorphic image 
 of the unit circle 
 into 
 such that every vertex (and hence every end) of 
 appears in 
. More details about these notions can be found in [D].
A graph 
 (finite or infinite) is said to be uniquely hamiltonian if it contains precisely one hamilton circle.
For finite graphs, the celebrated Sheehan's conjecture states that there are no 
-regular uniquely hamiltonian graphs for 
; this is known for all odd 
 and even 
. For infinite graphs this is false even for odd 
 (e.g. for the two-way infinite ladder), but each of the known counterexamples has at least 2 ends, leading to the problem stated.
Another way to extend Sheehan's conjecture to infinite graphs is to define degree of an end 
 to be the maximal number of disjoint rays in 
 and ask the following:
? Bibliography
[D] R. Diestel, Graph Theory, Third Edition, Springer, 2005.
*[M] Bojan Mohar, Problem of the Month
* indicates original appearance(s) of problem.
          
 Drupal
 CSI of Charles University