depth. However, Deutsch and Ramachandran's example of a crumpled newspaper clearly shows that this is not the only factor involved. In this case there is an abundance of high contrast local disparity information provided by the printed letters, yet the pseudoscopic image cannot be reversed in depth. This is also true if the words and letters are completely unintelligible. Likewise, pseudoscopic viewing of the normal face of our original figure 5 (1), upsidedown or partially covered, does not make the nose reverse. We therefore feel confident that the binocular interpretation of texture perspective (2) and not familiarity counter-