Identities for relations of the form (a,b). If it holds, how do you “prove” it?

First some context (binary relations)

$F \subseteq W \times X \quad G \subseteq X \times Y \quad H \subseteq Y \times Z$

Does the following equality hold

$(G \circ F)^{-1} = F^{-1} \circ G^{-1}$

I haven't been able to find a counterexample, so I think it holds in general. But how would I prove it?

This identity

$G \circ F = F^{-1} \circ G^{-1}$

I have a counterexample for.

And then the following identity

$(H \circ G) \circ F = H \circ (G \circ F)$

I have not found a counterexample so this could hold. But how would I prove it?

Edit: The task says give a good reason, so it doesn't have to be a mathematical proof.

-
proofwiki.org/wiki/Inverse_of_Composite_Relation proofwiki.org/wiki/Composition_of_Relations_Associative I would write the second proof differently (by rewriting the condition $(x,z)\in (H\circ G)\circ F$), but perhaps it's just a matter of taste. – Martin Sleziak May 1 '11 at 12:02
It would be helpful if you gave your definition of the composition and inverse. I would regard this as each binary relation being single steps and the composition being paths of two steps, so of course taking a path of two steps and reversing it is the same as taking reverse steps in the reverse order. – Phira May 1 '11 at 12:04