Underwater camera housings have a window through which the lens looks. When the camera surfaces, a couple of things can happen, and I've seen both captured in the camera footage. In one case, water clings to the window, resulting in a distorted or obscured image. In the other case, streams of water may briefly be seen flowing over the window, but this quickly subsides to allow a clear view. I assume the two different observations are due to surface characteristics of the window.
My question is: what accounts for the difference?
What role (if any) does surface tension play? What, at the molecular level, is producing two different results? Is it due to specific surface materials or treatments, or just the level of cleanliness and presence or absence of surface defects, scratches, etc.?