Many doubts have been cast on the transitivity of part; though, it seems that these doubts are based on 'local' problems, due to the various identity criteria an individual may assume. On the other hand, transitivity is not inherited from part to regional parts (see theory 'meronymy').
In general, regional part notions should be used; but when the axiom is sufficiently general, or when arguments lie in heterogeneous concepts, or when are regions, 'part' must be used. As far as regions are concerned, one could wonder which is the difference between located and part between regions. Thus, figure it out a geographic area A1 which is part of another area A2 without A1 being located at A2, for example so-called 'enclaves'.
(=> (Part-Of ?A ?B) (=> (And (Part-Of ?A ?B) (Part-Of ?B ?A)) (Identity ?A ?B)))