Adsorption of proteins from the bulk is at times accompanied by a rearrangement which leads to the formation of closed packed bodies, that may or may not be crystalline. Mass transfer of protein molecules on a surface is modeled. Forced diffusion by van der Waals and electrostatic forces leads to segregation, which is eventually a different phase that is assumed to be thermodynamically favored. The net effective force in two-dimensions has been modeled approximately and shown to be much stronger and more long ranged than in the bulk: that is, under the same conditions, the protein molecules may not aggregate in the bulk they may aggregate on a surface. These forces have been used only indirectly but equivalently as an adsorption-desorption step at the interline. Eventually, a linear stability analysis of the growing body shows it to be unstable and would give rise to whiskers that are one molecule thick. This is what is observed experimentally. The conditions that give rise to the instability have been determined. The reverse case of rinsing of the protein molecules has also been studied experimentally and has been analyzed using the same mechanisms. Here it is seen that thicker inroads into the packed body cause the interline to take on a spongy appearance. It is conjectured that eventually islands will appear as seen in the experiments.