Keith Snell and Paul Ell have written a superb and definitive study of the geography of religion in Britain circa 1851, relying upon a massive database constructed by them from the original returns of the 1851 census of religious worship, which the government undertook, for the first and only time, as part of its decennial population census. The religious census measured attendance at worship, not individual professions of faith; the returns were head-counts at morning, afternoon, and evening services on census Sunday, submitted by the presiding minister of each denomination. These returns, despite inevitable doubts about their reliability, have long been recognized by historians as an extremely valuable source. But until now they have not been analyzed comprehensively or with the aid of multivariate statistics. Not only do the authors describe the distribution of the strength of each denomination across the whole of England and Wales, using parish-, registration-district-, and county-level data; they are also able to address associated historiographical issues with a decisiveness that convincingly settles many an important debate.