Search Website Design and Content © by Eric Krause, Krause House Info-Research Solutions (© 1996)
      All Images © Parks Canada Except Where Noted Otherwise
Report/Rapport © Parks Canada / Parcs Canada  --- Report Assembly/Rapport de l'assemblée © Krause House Info-Research Solutions

Researching the Fortress of Louisbourg National Historic Site of Canada
  Recherche sur la Forteresse-de-Louisbourg Lieu historique national du Canada

DOMESTIC BUILDING CONSTRUCTION 
AT THE FORTRESS OF LOUISBOURG, 1713 - 1758

By

Eric Krause

1996 Draft Report

(Fortress of Louisbourg
Report Number H G 10)


CHAPTER SIX

ROOFS - ROOFING MATERIALS

There were six major roofing materials in Isle Royale: bark, wood, slabs, sod, shingles, slate and boards. Slate, however, was not used during the early years of the settlement.

The demand for stripped logs for piquet constructions and the stripping of live trees in the surrounding forest provided an inexpensive source of supply of bark as a roofing material. Damage was so rapid and pervasive in the latter case, however, that in 1717 an ordinance was issued forbidding outright the practice of live stripping and the use of bark as a roofing material.

The rising number of building starts might also explain the source and popularity of plans de bois, or wooden roof slabs, flat on one side and curved on the other, with the bark still attached. Builders choosing to square off their piquet logs (rather than using them in the round) were an obvious source. Another were sawyers and hewers of timber who would have produced a similar by-product.

Plans de terre, or slabs of sod with grass and plants attached, were generally reserved for the roofs of crude buildings. Their use was never widespread. Then, in 1717, the year that Louisbourg was chosen capital of Isle Royale, the king, having banned bark roofs entirely from the town, implicitly included sod and plan de bois in the ban too. Bark, sod, and wooden slab roofs were costly maintenance headaches and fire hazards, and accordingly boards and shingles were to be used for roofs until slate became readily available.

The technique of laying a bark or plan roof was little discussed. Wooden slabs were the base material for one sod roof; bark the base for another.

Wooden shingles eventually displaced both bark and plan materials, becoming the most popular roofing choice in the town. The memoir of 1717 merely confirmed that trend. Shingles already cost one-quarter less than bark as a roofing material and a supply was readily available, first from local splitters working during the autumn and winter, then from New England merchants. By 1751 the engineer Franquet was reporting that the New England shingle was totally dominant, having displaced the local product completely. Franquet, however, was not entirely correct. Periodic shortages were always possible, as in 1756, a time of war, when Louisbourg would place an order for Quebec shingles.

Several factors clouded the question whether there was a standardized approach to shingling at Louisbourg. Sources of supply were quite varied, though the New England shingle did dominate, and proposals that never advanced beyond the planning stage were numerous. Yet there were some common characteristics in all the descriptions: shingles were tapered with a reduced thickness from butt to head, they were placed in equal rows with the length of each shingle exposed no more than one-third, and, beginning in the 1730s they were nailed to a beveled board sheathing. The New England shingle, in particular, was pine, 18 pouces long, 4-5 pouces wide and 4 ligne thick, and nailed with two nails to a shingle. Other shingles varied in description: of oak or white fir, supplied in lengths of anywhere from 12 to 14 pouces; in widths ranging from 5-6 pouces to 9-10 pouces; with a butt of one pouce and a head of one ligne; and nailed either with three nails to oak laths of 4 pieds long, 3 pouces wide, 6 ligne thick, or to Boston boards that were nailed in turn with two nails to each rafter.

The beveled roofing board traveled the same evolutionary course, from Newfoundland to Louisbourg, as did the beveled weatherboard. There, at Plaisance, prior to the 1713 settling of Isle Royale, builders who were versed in the traditional use of laths had already begun to experiment with board sheathings and shingle coverings. This development was out of necessity as neither slate nor tile could hold in the wind.

In Louisbourg, the military, though committed to the use of slate, often chose shingles because of a problem in the supply of slate. Consequently, even the barracks of the King's Bastion was shingled for a time. But in those early years neither slate nor shingles were effective against driven water and powdered snow, and the elements were easily penetrating buildings through the spaced laths to which the slate or shingles were nailed. As early as 1716 then, several proposals were put forth for one pouce thick roofing board (of a type also suitable for floors and partitions) in place of laths. Undaunted, however, royal officials continued to use laths for some time. Between 1721 and 1725, for example, laths and wooden shingles - shingles were by now recognized as superior to slate, particularly in conjunction with laths - were placed on the roof of the commissaire-ordonnateur's residence. The use of laths, however, was drawing to an end, and during this transitional period, in 1723, engineer Verville would recommend boards with butt joints as a sheathing for the barracks roof of the proposed Royal Battery.

Finally, in the 1730s, the beveled roof board made its appearance. Always one pouce thick, in every other way it met the same specifications of the beveled weatherboard. A 1738 memoir also clarified its function: beveled roof boards were a second line of defence against the elements. A tight fit was critical, uniform beveling a necessity.

Beveled boards probably assured the continued use of slate on king's buildings, slate being particularly vulnerable to Louisbourg's climate. No less an example was the Block 1 artillery storehouse. Completed in 1737 the building leaked almost immediately, its slate roof (the type of sheathing is unknown) proving incapable of preventing water, snow or ice from entering. A serious situation indeed, it prompted a call for a replacement with shingles, but the military squashed the proposal, fearing an increased danger from fire.

In an effort to improve slated roofs the engineer, Verrier, conducted an experiment. He had the sheathing boards beneath the slates on a particularly leaky roof caulked with a torchis (or bousillage) mixture of clay and straw. His confidence in the procedure was unfounded, however. Not only did the caulking add weight to the roof, but it crumbled soon after it had dried.

Local slate, discovered two leagues from Port Toulouse in 1716, was of inferior quality and so Louisbourg, forced to use imported slate from France, looked towards Nantes, Angers and, in particular, to St. Malo, whose slate was considered to be the best. Roofing slates arrived in Louisbourg in pre-cut assorted widths of proportional lengths, in cases each containing from five to six hundred slates. An acceptable range of widths was 5-8 pouces (under 5 pouces was once declared unacceptable), although, inexplicably, a slate in the range of 5-7 pouces (with lengths proportional), described in a 1745 order for 26,000 slates, was thought to be large. A slate was dressed on three faces and nailed to the sheathing with two or three flat-headed nails, allowing for an average gauge of 3 1/2 to 4 pouces left exposed to the weather.

Board roofs, second to only shingled roofs in popularity, were poorly described. What is known is that those of Isle Royale were laid either in single or double layers placed horizontally or vertically to the eaves with butt, with board (spaced apart in one example) and batten, or with overlapping (clapboard style) joints as required or desired. Boston boards and thicknesses of one pouce or 1 1/4 pouces were acceptable.

Boards next to another material on the roof of the same building were not unknown. Accordingly, a piquet building at Louisbourg's "passage" had a roof partly in boards and partly in wooden slabs. Another, south of Louisbourg's barrachois, had one section of roof in bark only.

Back.gif (1009 bytes)
Return to the First Page/
Retour à la page première

     TO CHAPTER SEVEN     

Return to the Previous Page

Retour à la page précédente