Building on the base of evidence and interpretation in Hansen' s ( 1994 ) qualitative study of working people' s diaries, ,se assigned each diarist a set of codes to indicate employment, marital status, number of children, and size of the town in which he or she lived. To analyze the number, location and gender mix of visiting occasions, we coded each day in January and July for every year of the diary, counting the number of named visitors, the visitors' gender, the size of the visiting occasion (1 to 4 people, or 5 and above), the gender mix of those present during the visit, and the location of the visit. While this may seem straightforward at first glance, the variable nature of the diary entries meant that the coding process was not as uncomplicated as we initially anticipated. Given the number of diarists and the span of diary-keeping years, we faced the possibility of coding over 200,000 diary days. Because of the labor-intensive nature of the coding and the number of entries, we chose to code only 2 months—January and July—of each year a diarist kept a diary. We chose 2 months that could reflect a range of sociability. Severe January weather in New England impeded mobility, but it also freed those who were farmers from most of their labor-intensive chores. July tended to be haying season tbr farmers, which meant some people routinely worked all month in the fields—some alone, some with hired help. Further, the clement July weather meant grater mobility for all of the diary keepers. For some people—those who kept a diary for only a single year—the fact that we coded only 2 months out of each year meant we have only 62 'diary-days' to document their social lives. For others, we have several thousand. Limiting ourselves to January and July for each diary year, we nonetheless coded entries for a total of 24,752 diary days. In an effort to capture an accurate picture of visiting patterns, we coded every day of a given month, even those that had no entry or that mentioned only the weather, as well as those that recorded numerous visiting occasions in one day. Determining a working definition of what constituted a visit was also an unexpected challenge. For example, although schoolteacher Mary Mudge kept a meticulous record of her visiting 'rounds,' listing names, places, and conversation topics, other diarists were not as forthcoming. A typical entry in farmer John Campbell' s diary (9 July, 1825 ) was less amenable to our initial coding scheme: 'Go to Carr' s for Oxen.' ( See Hansen and Mcdonald, 1995, for a fuller discussion of the pitfalls of coding diary data. ) We therefore created the following coding protocol. We defined a visit as any occasion in which the diarist names the presence of individuals not of his or her household, the presence of the non-household member serving to distinguish between a community interaction and a household interaction. We also coded as visits public events at which the diarist was present but others in attendance were not named. The most common among these were records of church attendance. Although an entry 'went to church' did not result in a finding of specific male or female visitors, it was a community interaction thus, these entries were coded as gender-mixed visiting occasions of five or more people in a public place. Because of the variable nature of diary-keeping practices, we were careful to record only what we could confidently infer. Therefore, some entries record visits but no named individuals. Others, such as church attendance (which is generally a large-group event) or a visit to one named friend ( which is an intimate affair), allowed us to code the size of the group. Still others, when the location of the visit was specifically mentioned, allowed us to code the diarist as hosting, acting as a guest in another' s home, or interaction at a public place. What is the si