The census confounded most enumerators, and the resignation is evident in the above quote. Even if the 1851 and 1861 census enumerations could be considered the rookie years, the census was seemingly routinely “bungled” with each passing decade. In 1911, the Mayor of Fredericton complained about census results, demanding a recount. He suggested that they adopt the English approach to census enumeration, in which forms were left to be completed by the householder and later collected by the enumerator. It was suggested that this could improve accuracy, and that the census should be distributed by the Police Force, since they were the most intimately acquainted with the location of residences and their occupants. From 1851, New Brunswick followed the American model, with enumerators visiting individual households and recording the responses themselves. Clearly this approach had its shortcomings, and complaints about enumerators and census results were entirely too common.

Enumerators were generally well-connected, and were selected based upon the recommendation of the local MLA. For the 1851 census, Adam J. Beveridge, originally from Fossaway, Kinrosshire, Scotland, was the enumerator for Andover, Victoria County. He canvassed that large parish in 40 days, recording his own household, at the end of which he earned a salary of £20 for his efforts. Beveridge moved a few months later to St. Francis parish, where he was counted again! This is not the only case where a Scot was counted twice in the census. Donald Fraser was counted in both Glenelg and Newcastle, Northumberland County in 1851. Being a teacher, Fraser was recorded both at home and where he was keeping school. Alexander McRae, originally from Perthshire, was enumerated in both Hillsborough and Hopewell, Albert County in 1861. He was occupied as a Trader in Hillsborough and a Teacher in Hopewell, and different family members resided in his household in the 2 separate enumerations. Because mobility was fairly common, it can be assumed that there were even more examples of double enumerations.

Completed census schedules were returned to the Provincial Secretary’s Office at Fredericton, and different parishes and counties lagged behind the others. For those who were delinquent, it had been suggested that they be fined for excessive delays. For some, there were extenuating circumstances. The original enumerator appointed to canvass Westfield, Kings County, had fallen ill and was replaced with 3 new enumerators to expedite the process. However, one of the three enumerators took his time completing his portion of the parish. When the Clerk of the Peace visited the delinquent enumerator at home, he was working on his farm and complained of a sprained ankle. As soon as his schedule cleared, he promised to get back to work. His were the last forms turned in, and Westfield was finally complete in March 1852.

The process was no more improved a decade later when it came time to take the 1861 census. Enumerators expressed confusion about how to complete some of the columns, asking questions in letters sent to the Provincial Secretary’s Office such as whether it was best to write “Head of Family” or to leave that line blank. There was no best practice, no standard by which enumerators measured a successful census return. There appeared to be far more problems with completed census returns in 1861 than 1851, and a number of enumerators were later asked to consult with the Provincial Secretary and to provide clarification and make corrections to their completed forms. In the case of Norton, Kings County, the census was taken again, a month after it was originally completed. The second time around, a new enumerator was appointed. The results of the two enumerations, taken one month apart, showed some significant differences. A number of people aged 2 or more years in that month, and the household structure had changed with some family members leaving and new ones suddenly appearing. This re-take suggests that a census captures a particular moment in time, that change can be rapid, and that not all the information recorded can be considered completely accurate and reliable, that the responders themselves are fallible.

To improve the accuracy of information, an effort has been made not only to connect information across both the 1851 and 1861 census records, but to connect to other sources, including obituaries, biographies, and later census records. The census databases, then, are more than simple transcriptions, and we have done our best to ensure that these records are as useful and as complete as possible for all researchers.

There is no real way to compensate for missing records, and a number of counties and parishes for both 1851 and 1861 have unfortunately been lost. The forms, both in 1851 and 1861, had to be turned in to the Provincial Secretary’s Office at Fredericton, and compilers had been employed with the specific task of calculating census statistics. It must be assumed that once these tabulations were performed and the results submitted for publication in the Journals of the House of Assembly, that the paper copies of the census were returned to the counties. A number of these census records met an unknown fate, and how they came to be lost in whole or in part remains a mystery. In the case of Saint John, it has been suggested that the census records perished in the Great Fire of 1877. Curiously though, portions of both the 1851 and 1861 Saint John census have survived. There are no ready explanations, however, for why Gloucester, Kent, and most of Queens County are missing for 1851. Thanks to the work of the census compilers, though, there is a record of the population returns for the missing parishes and counties. Below is a chart documenting the number of Scots in every county for 1851 and 1861.