A key fossil in the study of how birds evolved from therapod dinosaurs, Rahonavis is a genus of dinosaur that lived in what is now modern Madagascar during the late Cretaceous period between about 70 and 65 milion years ago.
First discovered in 1995 and initially named Rahona, as is often the case in palaeontology it was subsequently discovered that an insect was already in possession of the name, specifically a genus of lymantriid moths. At first it was thought that Rahonavis was a new genus of bird, in particular because of the presence of quill knobs on its ulna (Although a few researcher have even suggested that the forearm limbs may not belong with the rest of the skeleton). However further study soon revealed that things were not so clearcut. It became apparent that in addition to its many bird like features, Rahonavis also shared many characteristics with dromaeosaurs (Raptor) dinosaurs, including a typical raptor-like raised sickle claw on the second toe. Ever since it has been the subject of quite a bit uncertainty as to its proper taxonomic position, the sort of uncertainty that results in lifelong feuds among palaeontologists.
Much of the reason for this uncertainty is that although the link between primitive birds and dromaeosaurids is well established, because there are close affinities between related cousins, and the possibility that flight may have developed and been lost multiple times among these groups, it has proven difficult difficult to place Rahonavis relative to the the lineage that leads to modern birds. Although it is not certain that Rahonavis could fly, the consensus appears to be that it would have been most capable of flight, if somewhat less elegantly than modern birds.
At about 50cm in length or so, Rahonavis was slightly larger than the most famous bird like dinosaur Archaeopteryx, roughly the size of a modern raven. Indeed was suggested by the original describers that Rahonavis could be a close relative to Archaeopteryx. However some pelvic adaptions suited for flight, while similar in function to those of Archaeopteryx, appear to have be independently derived. Most recent studies place Rahonavis in or as a sister taxon to the sub-family of dromaeosaurs called unenlagiines, and so more closely related to small dinosaurs like Unenlagia, Buitreraptor and Neuquenraptor. Interestingly if this classification is correct it suggests that either Unenlagia may have become secondarily flightless, having evolved from flying, Rahonavis-like ancestors, or else it is evidence that bird-like flight having evolved at least twice.