Much of the research for this has been carried out online. Wherever possible, demonstrably trustworthy and hopefully authoritative sources have been used: Journals such as Science, Nature and the sites of insitutions such as the Smithsonian. In other cases, and especially where references are to specimen or out-of-print sources, I've tried to make sure that corroboratory information is available from multiple sources, without their appearing to be direct copies of each other. Where Wikipedia contains a useful reference, I've tended to provide that here – as a collaboratively edited source, outlandish or inaccurate references tend to be edited out very quickly.