The law of superimposition says that stuff found on top is younger than stuff found lower down, in a geological or archaeological column. This is generally true, but there are exceptions, mostly trivial and easily understood. If a cave forms in a rock formation, the stuff that later ends up in that cave is younger in depositional age than the rock underneath which it rests (the rock in the roof of the cave, and above).
One of the coolest examples of what seems to be (but really is not) a violation of this Law of Geology is a thrust fault. A thrust fault is essentially a horizontal fault (as opposed to the more common vertical fault) in which rock from one area slides completely over another area. When this happens, the rock at the base of the upper unit (the one that slid over the other rock) is older than the rock on which it rests.
The fault isn’t really horizontal. but it’s horizontal enough for this to happen. And, in fact, the whole thrust-faulting thing is actually fairly complicated, and there are different processes that cause a similar effect. But in the end, you get an older layer sitting on top of a younger layer.
The reason this is not really a violation of superimposition is this: The older rock was actually deposited on top of the younger rock later in time than the formation of the younger rock. In a way, this is not much different than an ancient mountain made of ancient stuff eroding and generating sand that flows downstream and covers some pre-existing sediment. The fact that the grains of sand were formed a long time ago does not make that recently formed sand deposit old. It is young. But it is a young deposit made of old stuff. A thrust fault is the same thing but instead of there being a zillion tiny grains of sand deposited on some earlier sediment, it’s all one big giant piece!
Here’s a couple of photographs of the Keystone Thrust fault. Tell me if it looks familiar to you: