Comment on ‘‘Metal-insulator transition in random superconducting networks’’

We make some remarks on the determination of the normal-superconductor phase boundary in random superconductive networks. We discuss a recently reported work by Soukoulis, Grest, and Li [Phys. Rev. B 38, 12 000 (1988)] that introduces weak links between nodes as these are removed in the site-percolation problem. By the analysis of two simple geometries, it is shown that this procedure introduces spurious effects that mask the physical properties of the system. These affect, in particular, the field-slope critical index and the sharpness of the normal-superconductor boundary.