In picture 1, the Loft, a face emerges mysteriously from objects in the foreground and on the table. Its an ambiguous image, one that represents more than one thing within a single outline. The classic ambiguous image is the bird/rabbit figure shown to the left, often found in psychology text-books - its a bird facing right, but a rabbit facing left. Note that as it flips from one to the other, you can only see it one way round at a time.
The more elaborate example to the left below is adapted from a print made a century and a half ago, by Japanese artist Utagawa Kuniyoshi. Pictures of this kind, of heads made up of assemblies of creatures or objects, were made famous by Giuseppe Arcimboldo about four hundred years ago. To see lots of his pictures, just type Arcimboldo into Google Images.
However Arcimboldo did not invent this sort of thing. For a remarkable example from about fifty years before Arcimboldo got to work, take a look at the (rather rude) decoration on a plate that Oxford's Ashmolean Museum bought in 2003, at www.ashmoleanprints.com/image/383255/francesco-urbini-maiolica-plate-painted-with-composite-head
On Opticaloctopus, in comic book picture 6, the purr tree is another picture in this tradition.
The ambiguous face in comic book picture 1 is more complicated because the objects comprising it are distributed in depth within the scene. Ambiguous pictures of that kind first became common in popular books of puzzle pictures in the late nineteenth century.
If you want to study all sorts of ambiguous pictures in more detail, try the catalogue from a 2003 exhibition at the Museum Kunst Palast, Dusseldorff. The english version is: Dawn Ades, Jean Hubert Martin and Stephan Andreae, The Endless Enigma, published by Hatje Cantz and available through Amazon.
It's not specially cheap, but it's crammed with images from all over the world, including lots of examples by Salvador Dali.
For your own experiments, start with heads and faces. So much of our brains are specialised for perception of those that we discover them easily where they don't exist (as in clouds). But then take a look at David Lewis-Williams, The Mind in the Cave, Thames and Hudson, 2002, plates 17 and 18. It's an image from tens of thousands of years ago, in which the painted details added to a cave wall conjure up a bison, but only if the light from a lamp strikes the rock so that its contours appear in relief, as the bison's spine. So the effect is a flip between seeing the surface as rock and as bison. That suggests some mind-bending three-dimensional artwork.