There have been a lot of "mimics" or "false" versions of animals throughout evolution. Thylacosmilus was a marsupial saber-toothed tiger from South America. Janjucetus was a baleen whale from Australia who wanted very much to be a sea lion. Triassic henodonts were turtle-mimics. Of course, modern glass lizards mimic snakes very well (only their eyelids give them away). Well, as it turns out, one small group of archosaurs independantly evolved a dinosaurian bipedal posture, a beaked jaw, small forelimbs, and an eerily saurischian-style pelvis. What make these little beasts so wierd is that they're actually crocodilians. As I've said many times before, Archosauria is generally split into two major categories, Ornithodira and Crurotarsi. Specifically, these strange creatures are among the rauisuchian crurotarsians, a broad group that includes Postosuchus and Arizonasaurus. The group in question, which includes little Effigia okeeffeae (above), has no formal name thus far (although "Chatterjeeinae" has been batted around) but is rooted within the poposaurids, a small subgroup of the rauisuchians.
The first "Chatterjeeian" discovered was Shuvosaurus, and those remains were fragmentary. But among the fossils found, a toothless beak was noticed. Given the rest of the surprisingly dinosaurian bones, its discoverers thought that it was an ornithomimid, and given its Late Triassic age, a very primitive one. The implication here was that the majority of the theropodian clades have diverged by the Late Triassic, an idea not accepted by many. On this note, I should point out that Eshanosaurus, a "therizinosauroid" based on a Late Triassic mandible, is most likely a misidentified prosauropod.
Just recently, Effigia was discovered from an amazingly complete skeleton and skull, and has shown that the "Chatterjeeidae" is actually a brand of crocodilians who took covergent evolution to the extreme. How extreme? Here's Effigia's skeleton:
There are a few telling features that give away Effigia's crocodilian ancestry, including its "crocodile-normal" ankle (which allowed less motion than the ornithodiran model), plate-like scapula, and basic structure of the foot (among other things). But the pelvis is amazingly saurischian, including a pubic boot. The mouth is completely toothless and full of fenestrae. It's little wonder to me that, based on fragmentary remains, one would come to the conclusion that one of these little wonders was a dinosaur. In his description of Effigia's remains, Nesbitt lists the many covergences that the fossil shares with theropods. Incredibly, Effigia converges with not just the Dinosauria, but also the Neotheropoda, Coelurosauria, Ornithomimosauria, and finally, the "Edentulous Clade of Ornithomimids" (beaked ostrich dinosaurs). What's even more amazing is that Effigia is from New Mexico and shared territory with ceratosaurian dinosaur Coelophysis and more primitive rauisuchian Postosuchus.
It's crazy enough that a group of crocodilians managed to figure out the theropodian bauplan independantly of actual theropods, but for the "Chatterjeeidae" to zip forward over 100 million years to converge with the ornithomimids? That's mind-blowing. Effigia okeeffeae and its sister taxon, Shuvosaurus inexpectatus really speak to the plasticity of evolution, and offer up the greatest example I've ever seen of convergence.