The molecular shape of a network can be characterized in detail by using in
formation on its geometry and topology. In the case of permanent networks (
e.g., those formed by a polymer melt), the structural complexity is usually
measured in terms of entanglements within a chain and between chains. In t
his work, we show that this viewpoint can also be extended to the structura
l analysis of transient networks (e.g., those found in liquids with hydroge
n bonding). For the hydrogen-bonded networks within clusters of simulated l
iquid water, we show the existence of power-law scaling in the averaged ent
anglement complexity as a function of cluster size. (C) 2000 Published by E
lsevier Science B.V. All rights reserved.