There’s a great piece up at ScienceBlogs demonstrating how paleontology (and, even better, paleontological art) really works. It’s Darren Naish enthusing about a particularly strange-looking cousin of the allosaurs:
What interests me in particular about Concavenator it that it essentially vindicates my stupid suggestion: in this theropod, the 11th and 12th dorsal vertebrae have really tall neural spines, five times taller than the centra. The spine of the 10th vertebra is shorter, but it’s still taller than the preceding 9 dorsals. While I assumed – I would say quite reasonably – that Becklespinax had tall sacral neural spines, Concavenator has really short ones that don’t even make it over the dorsal edge of the ilium. So, Concavenator had a short, tall sail (or hump) that rose abruptly from the posterior part of the back, and was restricted to this part of the back alone. As is obvious from Raul Martín’s excellent life restoration [used at very top], it would have made the animal look like it had a dorsal fin….
(Emphasis mine, so you know what he’s getting at. This excerpt might be hard to follow if you haven’t read what came before it, but it captures the way Naish gets excited by one discovery lending support to another – the way the pieces join together.)