My favorite theropod when I was growing up, Eustreptospondylus oxoniensis (from the Oxford Clay, a mid-Jurassic English near-marine deposit) was at first the most svelte, and at second the most elegant mid-sized theropod I could find. Greg Paul's reconstruction and art made it seem all the more so. Facts, however, tend to put dampers on romantic views of old specimens. You see, the animal wasn't fully grown, and the proportions of the limbs to the shoulder and pelvic girdles, and unfused vertebrae, suggested it wasn't even close to adult size. There is, of course, few ways to determine exactly how old the specimen is, and no one has had the nerve to cut into the animals femur, tibia, or ribs to count the lines of arrested growth and measure the medullary cavity to check.
That all said .. Eustrepto was considered an allosaur relative by many, with vague hints at megalosaurs. But new data suggests that the megalosaur relationship is more likely to be true, and indeed Eustrepto seems to be a basal megalosauroid not that far removed from the spinosaurs. More interesting is that these animals had relatively small pelvises and huge heads for their size, facts especially ture of close-relative Torvosaurus tanneri. So the seeming proportions as a young animal are somewhat erroneous, and this animal may be close to adult size.
This skeleton reconstruction is based on a single specimen, OUM J13558, which is not complete. Much of the head is present, and the vertebral column, but of the arm and shoulder only the scapula and humerus are present; and of the hip the pubis is missing. These elements were modified from other megalosauroids, and they help demonstrate how very lanky this animal is, but also emphasize how juvenile it seems.
Prepared for an upcoming book on the dinosaurs of the British Isles.