When Leonardo painted the image of Mona Lisa, he deliberated painted the edges of her smile in a slightly blurred method known as sfumato while leaving the rest of her mouth in sharp focus. The end result is meant to play with our eyes and the differences between our peripheral vision and central vision within the retina. While the retina allows us to focus directly on objects of interest and thus see them in very high detail, our peripheral vision developed as a way for us to take a general scope of our surroundings. Able to pick up slight movements and changes in our environment as well as general shapes, peripheral vision alerts our brain as to where we should look next but doesn't provide any great focusing ability. By painting the center of Mona Lisa's mouth in clear detail but leaving the edges fuzzy, Leonardo effectively prevents us from seeing the smile when we view the face head on (using our central retinal vision) but allows it to sneak back in when we focus on some other aspect of the piece thus allowing our peripheral vision to take over. Pretty sneaky Leo....
It's too bad that I'm not a professional neuroscientist. These findings were just presented in October of last year at the Society for Neuroscience's annual meeting in Chicago, and it would have been interesting to go. Click here for more information.