The list in Table A.1 shows the date when a city uniformed its police department and the rank size of the city (relative to all cities with information) in the next census year after adoption. It may not be apparent, but this list is the result of a good deal of time-consuming effort, particularly on the part of well over 100 librarians and archivists. Those cities for which I have cited no date belong to the frustratingly large category with missing information. I presume that the information is randomly missing and that complete information would fill out the plotted curves more adequately. Although the dates for a few cities were easy to determine, most required the energy of many people: For the cities where I have not reported any dates, I have been assisted by both librarians and historical societies in attempting to determine the precise moment at which the police were uniformed. If the missing information has a nonrandom structure, two possible kinds of bias could be introduced. Most likely could be the situation that places without noted dates were late adopters, the adoption of uniforms being commonplace and nonnoteworthy by the time of the late adoption. This could change the shape of the diffusion curve from an S to a U, the late adopters coming as a burst in the 1880s and 1890s.