Michael:
 > & since this was probably the most difficult such maze ever found 
> anywhere in nature
> you have practically proved that what i have been calling border 
> crosses are not necessarily border crosses at all
> & tho you havent proved they dont cross
 
Somewhere out there I imagine this simple situation of directed line 
segments, quadripoints and an Euler tour has been addressed in the 
maths literature, but I wouldn't know where to begin. Where is Martin 
Gardner's or Ian Stewart's column in Scientific American when you 
need it? Ian Stewart in particular might have picked this up and run 
with it.
What we can say, though, is that a *fully crossed* hotwater tour is 
impossible - the tiny dangling exclave attached to a larger clave by 
two quadripoints prevents it, since the only fully crossed walk 
around this would require two trips - one crosswise and one closed 
loop. 
A fully crossed tour was always going to be harder, since you only 
have *one* way out of any given quadripoint (straight ahead), whereas 
I'm allowed two (left or right) and can therefore better avoid being 
forced back to my starting point or creating detached, untravelled 
loops (which are the only two rules of Fleury's algorithm).
 
 > so er what would you term these
 
Are they not just bipartite quadripoints?
Grant