284 Evolution? The Fossils Say YES!
archosaurs as well, especially pterosaurs). Feathers, then, did not evolve for flight but were
already present in theropod dinosaurs, presumably for insulation, and were later modified
to become flying structures.
Prum and Brush (2003) have completely rethought the origin of feathers and showed
that they are not modified scales (as once believed) but from a similar embryonic primor-
dium with different Hox genes controlling development. Type 1 feathers (fig. 12.13) are
simple hollow pointed shafts, which appear in the primitive theropod Sinosauropteryx.
Type 2 feathers are simple down with no vanes, and type 3 feathers have a vane and shaft,
but no barbules linking them together like Velcro. Both of these types are found in the
large therizinosaur Beipiaosaurus, suggesting that they were present in almost all theropods
(fig. 12.13). Type 4 feathers have barbules that link the vanes of the feather into a continu-
ous surface, but the shaft is symmetrically aligned down the middle of the feather. This
kind of feather appears in Caudipteryx, which suggests that these occurred in higher thero-
pods (including Tyrannosaurus rex) as well. The classic asymmetric flight feather with the
shaft near the leading edge of the vane first appears in Archaeopteryx, and for this reason
many scientists think that Archaeopteryx was one of the first to modify the long heritage of
feathers for true flight.
Moving up from Archaeopteryx on the cladogram of birds (fig. 12.12), we come to Raho-
navis from the Cretaceous of Madagascar (Forster et al. 1998). About the size of a crow
FIGURE 12.12. The family tree of Mesozoic birds, emphasizing some of the recent fossil discoveries. (Courtesy
L. Chiappe)