The nineteenth-century city, and most especially the Canadian city, provides a rich arena for quantitative studies of literacy. The key is the availability of sources for such a project - for the manuscript census remains the best starting-point for the systematic study of literacy. Researchers who have used these schedules have ignored the literacy data, often dismissing it without consideration. This has resulted in a lost dimension in their work for these schedules provide the easiest, most direct, and perhaps the best information on the problem. The manuscript census furnishes the researcher with a relatively complete and unbiased roster of the people of a given geographical region, differentiating between those who could and could not read or write. Secondly, it provides a tremendous amount of direct information - demographic and economic - on each head of household as well as the rest of the members. It also serves as a standard measurement, as the same question was asked of each respondent. These schedules are useful for the United States from 1840 and for Canada from 1861, while the Canadian urban census of 1851 permits analysis on the basis of signatures for heads of households.