Nature © Macmillan Publishers Ltd 1998 8 that aid blood-feeding 8 . By so doing, they also promote the transmission of arthro- pod-borne pathogens 10,11 . Hui Wang, Guido C. Paesen, Patricia A. Nuttall NERC Institute of Virology & Environmental Microbiology, Mansfield Road, Oxford, OX1 3SR, UK e-mail: pan@mail.nerc-oxford.ac.uk Alan G. Barbour Departments of Microbiology & Molecular Genetics and Medicine, University of California, Irvine, California 92697-4025, USA 1. Wang, H. & Nuttall, P.A. Parasitology 111, 161–165 (1995). 2. Wang, H. & Nuttall, P.A. Parasite Immunol. 17, 517–524 (1995). 3. Wang, H. & Nuttall, P.A. Parasitology 109, 525–530 (1994). 4. Sauer, J. R., McSwain, J. L. & Essenberg, R. C. Int. J. Parasitol. 24, 33–52 (1994). 5. Willadsen, P., Eisemann, C. H. & Tellam, R.L. Parasitol. Today 9, 132–135 (1993). 6. Rigat, B. et al. Biochemistry 36, 8325–8391 (1997). 7. Wang, H. thesis, Univ. Oxford (1995). 8. Wikel, S. K. The Immunology of Host–Ectoparasite Arthropod Relationships (CAB International, Wallingford, 1996). 9. Alcock, J. Annu. Rev. Entomol. 39, 1–21 (1994). 10.Jones, L. D. et al. Science 237, 775–777 (1987) 11.Titus, R. & Ribeiro, J. Science 239, 1306–1308 (1988). scientific correspondence 754 NATURE | VOL 391 | 19 FEBRUARY 1998 raises the possibility that the Velociraptor and theropod ‘furculae’ are nonhomologous to that of birds — the view of Bryant and Rus- sell 2 . Fig. 1 of ref. 1 does not show an avian articulation of the ‘wishbone’ to the rest of the shoulder girdle and resembles the restoration of Barsbold for Oviraptor and the mounted specimen of Ingenia. Third, in most flightless birds the furcula degenerates into two clavicular splints 6 , simi- lar to the clavicles reported in some dino- saurs 2 . This strongly indicates that flight arose as an original function in birds, not as a modification of a structure already present before flight evolved 1 . Fourth, the presence of a furcula-like structure in Longisquama 6 , a primitive Trias- sic archosaur with long feather-like scales and postulated arboreal habits, indicates that structures of this type have developed more than once in the archosauromorphs, which means they are weak evidence for a bird link to theropods. Alan Feduccia Department of Biology, University of North Carolina, Chapel Hill, North Carolina 27514, USA Larry D. Martin Museum of Natural History, University of Kansas, Lawrence, Kansas 66045, USA Norell et al. reply — First, maniraptoran theropods, and theropods with furculae, are known from the Late Jurassic period 7,8 , sig- nificantly closing the ‘gap’ discussed above. We did not say in our Scientific Correspon- dence 1 that Velociraptor was a direct ancestor, only that dromaeosaurs are related to birds. How do Feduccia and Martin explain the presence of primitive living mammals such as monotremes and marsupials 100 million years after they split from the line leading to ourselves? Second, exact similarity is not a require- ment for homology — take, for example, the forelimb variation among the living orders of mammals. Bat wings, whale flippers and human arms are very different but homolo- gous. The furculae of living birds are highly variable 6 and often different from the furcula of Archaeopteryx, yet Feduccia and Martin do not question the homology between these two groups. These authors identify this ele- ment as a possible interclavicle, a bone that is extremely different in crocodylomorphs and not present in any ornithodiran 9 . We do not think Feduccia and Martin can, from a small photograph, say with con- fidence that the furcula is articulated along the entire margin of the coracoid. Examina- tion of the actual specimen shows that the furcula attaches to the scapula just as it does in modern birds. We discussed this feature in our Scientific Correspondence 1 : “…the proximal process of the furcula tapers to a point where it contacts the scapulocoracoid”. Third, we disagree with Feduccia and Martin’s implication that preconceived process dictates the interpretation of pattern. Science is about discovery; by assuming less, we discover more. Feduccia and Martin believe that furculae cannot exist in the non- avian Maniraptora because they did not fly, but this view precludes the possibility of fur- culae ever being discovered in these animals. Fourth, Longisquama is a poor-quality fossil, and the interpretation of single ele- ments is controversial. Longisquama lacks other characters — present in non-avian Maniraptora — that would ally it with birds. Single features do not overturn a hypothesis that is strongly supported by a plethora of character evidence. Ironically, if one does use Feduccia and Martin’s reasoning that Longisquama is a close bird ‘ancestor’ as advocated elsewhere 10 , the temporal paradox increases. Longisqua- ma comes from rocks about 220 million years old, creating a fossil-free gap of more than 80 million years before the appearance of Archaeopteryx. Any empirical measure of stratigraphic fit 11 will prefer a hypothesis of maniraptoran relationships over this one. Mark A. Norell, Peter Makovicky Department of Vertebreat Paleotology, American Museum of Natural History New York, New York 10024, USA James A. Clark Department of Biological Sciences, George Washington Unviersity, Washington, DC 20052, USA 1. Norell, M. A., Makovicky, P. & Clark, J. M. Nature 389, 447 (1997). 2. Bryant, H. N. & Russell, A. P. J. Vert. Paleont. 13, 171–184 (1993). 3. Thulborn, R. A. Zool. J. Linn. Soc. London 82,119–158 (1984). 4. Martin, L. D. Courier Forschungsinstitut Senckenberg 181, 23–36 (1995). 5. Romer, A. S. Osteology of the Reptiles (Univ. Chicago Press, Chicago, 1956). 6. Feduccia, A. The Origin and Evolution of Birds (Yale Univ. Press, New Haven, 1996). 7. Gauthier, J. A. Calif. Acad. Sci. Mem. 8, 1–55 (1986). 8. Makovicky, P. J. & Currie, P. J. J. Vert. Paleont. (in the press). 9. Sereno, P. C. J. Vert. Paleont. 11 suppl. 4, 1–53 (1991). 10. Feduccia, A. & Wild, R. Naturwissenschaften 80, 564–566 (1993). 11. Siddall, M. E. J. Vert. Paleont. suppl. 17, 76A (1997). Norell et al. 1 describe a Velociraptor ‘wish- bone’ which they interpret as a new piece of evidence for the theropod origin of birds. The bone fits a pattern of furcula-like struc- tures that have been discovered in certain late Cretaceous theropods, including Oviraptor, Ingenia and possibly others 2 , some of which possess clavicles — usually not fused. But this interpretation gives rise to problems of chronology, structure and function. First there is the temporal paradox: the earliest-known bird, Archaeopteryx, occurred 75–80 million years earlier than the Veloci- raptor and other late Cretaceous theropods that had ‘wishbones’. Supposed Jurassic examples are from carnosaurs, a group gen- erally thought to be distant from birds — although the cladistic analysis of Thulborn 3 places Tyrannosaurus closer to modern birds than to Archaeopteryx. Second, the furcula-like structure in the specimen of Velociraptor has a round cross-section 1 , while that of the primitive Mesozoic birds — Archaeopteryx, Confuciu- sornis and the enantiornithines — is dissimi- lar, being grooved postero-dorsally along almost its entire length 4 . Also, the articulation of the arms of Velociraptor’s furcula-like structure along the entire margin of the coracoid (Fig. 1 of ref. 1) is unlike the articular relationship of the fur- cula in birds. However, it is similar to the relationship of the interclavicle bone to the coracoids found in primitive diapsids 5 . This Theropod–bird link reconsidered Using a large human male population of 507,125, we find clear evidence for a depen- dence of body height at age 18 on birth month. Over 10 years there is a sinusoidal variation with a period of 1.0 year with maxima in spring and minima in autumn differing by 0.6 cm. Although global envi- ronmental factors 1 are small and can be studied only with the help of sophisticated methods on very large sample sizes, they might offer insights into still undiscovered mechanisms of human development. This may provide empirical facts for clinical Height depends on month of birth