Library and Archives Canada Materials

Processing the War Loss Claims

The primary source materials for this project are the documents submitted by inhabitants of Upper Canada to the British government to seek recompense for losses they incurred during the War of 1812. The process through which inhabitants submitted their claims is described in the Introduction and in Module 4. This section is focused on the processing of the compiled records from 200-year-old paper into digital formats and data sets. The records of the Boards of Claims for Losses are currently held by Library and Archives Canada in a collection entitled “The War of 1812: Board of Claims for Losses, 1813-1848, RG 19 E5A.” The collection was originally organized into volumes and files containing:

• Registers
• Administrative records of the Boards hearing claims
• Indexes, registers, and schedules of claims
• Claims
• Vouchers for payments, with schedules
• Duplicates
• Certificates and affidavits
• Powers of Attorney
• Register of receipts of payments

In the 1980s, the entire collection was microfilmed to provide researchers access to the material while protecting the fragile originals. In that process, 12.5 meters of textual records were translated into twenty-three microfilm reels and archivists added typewritten indexes, registers, and cover pages for each volume. These supplementary materials help researchers understand and explore the idiosyncratic arrangement of the original materials.

At some point before 2009, staff at the LAC digitized the twenty-three reels of microfilm and made them available to the public online. However, the digital files remain in the order and size prescribed by the microfilm reels. Each digitized reel on the archived web page contains up to 1400 individual files presented in an outdated viewer that loads each page of the collection as a separate PDF file. Because reels of microfilm were not coordinated to the contents of the original volumes, each digitized reel begins and ends on random pages. Although the indexes to the volumes are enormously valuable, they provide little help in navigating the digital microfilm reels. Ironically, searching the digital collection in its archived web page for claims and information pertaining to specific individuals is more difficult and time-consuming than using the original paper documents, where the index pages would speed up the process.

Screenshot of the Library and Archives Canada archived web page with a list of available digitized microfilm, last modified in November 2009. Visit webpage.

Screenshot of the Library and Archives Canada archived web page showing a single digitized microfilm page as a PDF and navigation buttons

To make these materials easier to search, read, and present, I developed a process to download all the digital files made available through the LAC web page, recombine them into complete microfilm reels, and then separate and collate them according to the original structured system of volumes and files. This process converted 20,000-30,000 individual files into 23 files based on microfilm reel into 158 files based on the volume/file structure. The newly recombined PDF files have been submitted to Library and Archives Canada and are now available through their newer digital collections site (which is improved but still difficult to navigate).

Digital Reconstruction Process

1. For each digitized microfilm reel, identify the URL of the first PDF document. Example: [url]/t-1122-00001.pdf
2. Identify the total number of pages in that reel. Example: t-1122 has 911 pages
3. Using the python script below, generate a text document with URLs for each page in that reel. Example: [url]/t-1122-00001.pdf, [url]/t-1122-00002.pdf, [url]/t-1122-00003.pdf, etc.
4. Using Automator for MacOS, download each PDF into a single folder for that reel [screenshot of Automator workflow]
5. Using Adobe Acrobat, combine PDFs into a single PDF document of 900-1300 pages.
6. Manually separate reel PDFs based on original volume and folder structure. This stage was greatly aided by the index and title cards that archivists inserted into the microfilm to identify break points between the volumes and files.
7. Rename PDFs using naming convention derived from original structure. Example: RG19E5-V3728-F1.pdf

Example of Process


Input: Microfilm T-1122 - 911 separate PDF files

Output: 22 files separated into:

• Index
• Register
• Volume 3728 Files 1 to 6
• Volume 3729 Files 1 to 12
• Volume 3730 Files 1 and part of 2

Python Script

    # This script creates a list of URLs
    # to download each page of the PDFs
    # for each reel available on the Library and Archives Canada website

    # Set start point to first page of PDF
    n = 1

    # Open temp file to store URLs
    f = open("temp.txt", "w")

    # Set upper limit to last page number of PDF
    # can be lowered to create smaller batches
    while(0 < n <= 6972):
    
    # Set to url of PDF without the page number
    url = "[URL for LAC content]"
    
    # adds the pdf file type to URL
    suffix = ".pdf"
    
    # extends page number to five digits
    page = str(n).zfill(5)
    
    # write each URL to the file
    f.write(url + page + suffix + "\n")
    
    # increase page number by 1 until limit reached
    n = n + 1

    # Close file
    f.close()

Identification and Transcription

Using the newly collated RG15 E5 collection, I identified claims pertinent to my research by reading through various lists of claimants, including three volumes of registers (Register 1, V4357, and V4358), a schedule of payments (V3737-F3), and indexes to primary and supplementary vouchers (V3760-F1, V3763-F2, and V3763-F3). I looked for any names that could represent a female claimant, erring on the side of inclusion rather than relying only on typically female names. Using this method, I identified 167 claims that were either submitted by women or were later managed by women who were widows or executors of male claimants who died before the final payments were made. These claims represent 8.1% of the 2055 claims reviewed by the board of commissioners in 1823. The women represented in these claims were largely white Anglo-American women, but also includes three Black women and one Native woman. The women also represent a diverse cross-section of economic and social ranks in Upper Canada. Properties ranged from small owned or rented farms and basic dwellings in town to large estates with extensive land in the township. Women from subsistence farming families with little social standing submitted claims alongside women from the most influential families in the province. While reading and transcribing these claims, I found that most could be grouped into two categories.³⁴¹

The first category includes claims made by women directly. In most cases, these women were either unmarried or widowed before making their claim to the board. Jemima Stewart submitted a claim for her losses that mentioned her husband Alexander, who had died during the war.³⁴⁰ Only a few claims were submitted by married women whose husbands were still living at the time of submission, as in the case of Lydia and John Evans, a couple who operated a tavern in York and submitted a claim together in 1816.³⁴¹ Lydia continued to operate the tavern while John was a prisoner of war and signed the claim documents in her own hand, which may indicate that she was both literate and active in their business and legal matters. Another couple, Catherine and Thomas Pool, submitted a joint claim in 1823 that combined two separate claims to an earlier board for losses that they incurred before they married.³⁴² In another unusual case, Sarah Clark (unmarried) submitted a claim with her brother George, as they were apparently co-owners of the property that was lost during the war.³⁴³

The second category includes claims administered or re-submitted by widows or female executors of men who had previously submitted claims but died before the bureaucratic process was completed. These cases most often involve a claim submitted to the first board of commissioners in 1815 and a subsequent claim to the board convened in 1823. For instance, Captain George B. Hall of the Provincial Navy submitted a claim in 1815 for damages to his home and storehouse for his business. Hall died in 1821, leaving his wife Angelique to maintain their family and seek compensation from the government. Angelique petitioned the 1823 board of commissioners for a copy of the original claim and re-submitted a new claim.³⁴⁴

Once I had identified the pertinent claims, I began to transcribe the documents in each case. Although most claims are between three and four pages, some are much longer. Elizabeth Thompson’s 21-page claim detailed a dispute between she and her son-in-law over property lost in the war. Her claim was produced in collaboration with four justices of the peace and three deponents attesting to Thompson’s ownership and loss of the property. Thompson’s claim is over 1,500 words, which is on the high end for a single claim. In total, the transcribed claims include over 60,000 words attesting to property lost, recording witness testimony, and appealing to the board of commissioners for adequate compensation.

Creating Datasets

Having transcribed the claim documents, I began to extract pertinent details from the claims to produce datasets to aid analysis. For each claim, I identified the following information where available:

• Original claimant name
• Final administrator or recipient of payments on claim
• Township of origin
• District of origin
• Initial valuation of losses
• Commissioners’ valuation of losses and amount awarded
• Individuals who provided supporting evidence for the claim, including witness statements and valuations of lost articles

This information was used to produce a dataset with points and links (nodes and edges) between claimants and individuals who provided supporting evidence. This allowed me to create a network diagram of the connections between community members as represented in the claim records.

I also cross-referenced the claims with registers and paylists to identify the amounts paid out in the initial round of payments by voucher (typically 25% of the awarded amount) and supplementary payments (typically 25% for claims unpaid in the first round or an additional 10% of the awarded amount). The extracted financial data allowed me to produce a dataset that allows calculations of totals for amounts claimed, awarded, and paid. These figures aid analysis of the cost of war losses to individuals and in aggregate, comparisons between different categories of claimants, and detailed figures for the influx of capital into the district as payments were made. Because the figures are represented in British pounds, shillings, and pence, I wrote custom formulae in Excel to calculate totals based on the non-decimal system in which 1 pound = 20 shillings and 1 shilling = 12 pence. Halfpennies (0.5 pence) and farthings (0.25 pence) are represented as decimal pence for ease of calculation.

Example: the following formula calculates totals for three different columns (£/s/d), then uses divisors and remainders to re-allocate the totals for shillings and pence into the correct larger unit of money. For instance, £24067 789s 230d is expressed improperly; this formula converts it to £24125 9s 2d (expressed as 24125/9/2 in a single cell).

=CONCATENATE(SUM(SUM(K2:K90),QUOTIENT(SUM(L2:L90),20),
QUOTIENT(SUM(M2:M90),12)),"/",
SUM(SUM(L2:L90)-(QUOTIENT(SUM(L2:L90),20)*20)),"/",
SUM(SUM(M2:M90)-(QUOTIENT(SUM(M2:M90),12)*12)))