Colour connections in e(+)e(-) annihilation

We have very limited knowledge about how the confinement mechanism works in processes like e(+)e(-) --> q(q) over barg$...g, when there are identical colour charges. In this case the partons can be colour connected by a string or a cluster chain in several different ways. We do not know if in such a situation Nature chooses a particular configuration at random, or if some configuration is dynamically favoured. Also in the perturbative parton cascade we have no well-founded recipe for describing the interference effects, which correspond to non-planar diagrams and are caused by identical colour charges. We have studied two different models, and are in particular interested in the possibility that a colour singlet gluon system hadronizes isolated from the remainder of the state. Using double tagged events with heavy <c(c)over bar> or <(b)over bar b> quarks, it appears to be possible to find a significant signal, if such events appear in Nature. If this type of (re)connected states appear in Z decays it may also be an indication that reconnection might appear between the decay products of two W's in the reaction e(+)e(-) --> W+W- --> q(1)<(q)over bar (2)>Q(1)<(Q)over bar (2)> at LEP2. This would be important, e.g. for a precision measurement of the W mass. (C) 1997 Elsevier Science B.V.