Using weekly mortality data for London spanning 1866-1965, we analyze the changing relationship between temperature and mortality as the city developed. Our results show that both warm and cold weeks were associated with elevated mortality in the late 19th-century, but heat effects, due mainly...