Abstract

John Nelder, who died in 2010, made various contributions to ornithology. The most dramatic was the statistical analysis of the Hastings Rarities, which demonstrated beyond doubt that some of them could not be genuine. It would have been formally impossible to prove that there were too many rare birds recorded in such a short span of years in such a limited area for all the records to be genuine; accordingly, he made no attempt to do so. Instead, with a fine display of forensic statistics, he showed that various other characteristics of the records were so unlikely to have occurred were they entirely genuine that it was unreasonable to accept that they were.The few criticisms of his analysis that were subsequently published were ill-informed and his work remains untarnished. It is possible to estimate a minimum figure for the proportion of the records that were false from the data that he presented; it is 60%.