A generic property of a first-order phase transition in equilibrium, and in
the limit of large entropy per unit of conserved charge, is the smallness
of the isentropic speed of sound in the "mixed phase". A specific predictio
n is that this should lead to a non-isotropic momentum distribution of nucl
eons in the reaction plane (for energies similar to 40 A GeV in our model c
alculation). On the other hand, we show that from present effective theorie
s for low-energy QCD one does not expect the thermal transition rate betwee
n various states of the effective potential to be much larger than the expa
nsion rate, questioning the applicability of the idealized Maxwell/Gibbs co
nstruction. Experimental data could soon provide essential information on t
he dynamics of the phase transition.