Canadian Historic Sites: Occasional Papers in Archaeology and History No. 6

The Excavation and Historical Identification of Rocky Mountain House

by William C. Noble

Interpretation and Historical Identification

With the excavation, description and analysis of Rocky Mountain House completed, it now remains to synthesize the amassed data into a correct and meaningful history of the fort. This task is by no means without problems, one of the foremost of which is the coordination of historical documentation with the archaeological evidence. To a large degree the goals of archaeological and historical research are combined in this task.

The following interpretations, however, are based primarily on archaeology. This is done through necessity as much as by intent. No clear description or continuous sequence of events can be obtained for the fort from the historical records. The documents are just too few, ambiguous and even untrustworthy in some cases. Nevertheless, important information can be gleaned from the documents to aid in establishing the time periods for some of the major architectural changes in the fort. The integration of these written and excavated details forms an important premise for any correct interpretation and historical identification of the fort.

Rocky Mountain House (FcPr-1) definitely pre-dates the later version (FcPr-2) built and operated down-river by the Hudson's Bay Company from 1866 to 1875. FcPr-1 also marks the site of a former Hudson's Bay Company fort as well as an earlier structure. The identification of these two forts at FcPr-1 constitutes the main problem discussed in this chapter.

Two Forts on the Same Site

There is no doubt that two different forts existed on the same site of Rocky Mountain House. This important fact is fundamental in clarifying interpretations relevant to the formulation of correct historical identifications.

As described in an earlier chapter, the primary evidence for the dual occupancy of the site lies in the trench feature. This trench, cutting across the interior southern end of the fort, represents the position of an original southern wall of exterior palisade. When this wall was standing the fort measured 100 ft. long north-south by 90 ft. wide.

At a later date, the original fort was enlarged and renovated. The original southern wall in the trench feature was dismantled and a 16-ft. extension was added to the fort's southern end. New buildings were also erected during the enlargement and, in effect, a new fort emerged measuring 116 ft. long north-south by 90 ft. wide. Let us now consider each of these two forts in detail.

The Original Fort

Details about the original fort are meagre. As noted above, the original extent of the fort can be determined as covering an area of 9,000 sq. ft. Accompanying and fortifying the exterior palisade walls, which are believed to have ranged from 12 ft. to 15 ft. high, were two bastions. One was positioned at the north corner of the fort and the other sat over a gateway in the south wall of the exterior palisade.

The north corner bastion, 13 ft. east-west by 11 ft. north-south, was built of vertical pickets, commonly referred to as en pile architecture (Jefferies 1939: 375). This is an early French style and is not characteristic of the Hudson's Bay Company. The occurrence in this bastion of interior corner support posts suggests that it had a superstructure. The second bastion, 8 ft. square, was one of overhead construction supported on four large corner piles spaced 8 ft. apart. All that remained of this structure were the four corner post pits marking its original position. A southern gateway existed under this bastion and, as a composite feature, both structures were retained in the second enlarged version of the fort. In both of these forts the south entrance was the main gateway.

Five features are all that can be reconstructed in the interior of the original fort. This is due to the fact that almost all of the original architecture was dismantled after the enlargement of the fort; it is also a reflection of the very nature of the early style of architecture as explained below. Building II? appears to be the one surviving building of the original fort. Its simple architectural style is one in which the foundation joists and sills were placed directly on the ground without sunken support piles. This is definitely uncharacteristic of the Hudson's Bay Company, and is considered by Garth (1947: 219) to be probably typical of North West Company construction. If ground-laid foundation joists were removed, little evidence would remain to mark the location of a former building of this architectural style.

Four pits also appear to be associated with the original fort. These are pits 10, 11, 12 and 18 in the northeastern area of the fort. Pits 10 and 18 are large cellar-like depressions with distinctive beavertail step-down entrances. This characteristic entrance only occurs at these two pits and is unlike the construction of known Hudson's Bay Company pits on the site. Two buried humus layers interpreted as representing two former sod horizons are present in pits 10 and 12. This evidence and the sequence of fill in each would indicate an early date for these pits. It is also pertinent to point out that no building remains are preserved over the cellar-like pits 10 and 18. Possibly building I once stood over pit 18, and was simply moved westward for subsequent attachment to building II. Pit 11, housing the subterranean building VII interpreted as a cold-storage repository, also appears to be early. The nails from this pit date to 1810-25.

Artifacts of early 1810-20 date are conclusively represented at the site. Some have proveniences in the above-mentioned features of the original fort. In particular, glass seed beads show a close association with early features; of a total 151 specimens recovered, 139 come from building I and pit 10. Similarly, many of the nails from the fort are early hand-wrought types or early cut nails dating between 1800 and 1825. The two wrought csk. nails marked "special" by Dr. Dove date to 1800 and come from building I. Other early artifacts are the seven bail fasteners of North West Company style. These constitute the single most diagnostic and identifiable artifacts pertaining to that company. Their distribution over the site is random, but there is a negative association of copper specimens with Hudson's Bay Company structures. One copper bail fastener was found in building I.

Other artifacts associated with features dating prior to the enlargement of the fort include 1 hawk bell; 1 brass pin; brass buttons of type G and D style; 1 sherd of black basalt stoneware; 1 sherd of grey salt-glazed stoneware; 1 clasp-knife blade bearing the stamped letter D, and 1 small iron keg hoop. Of probable or possible early date are an additional three pieces of undecorated lead-glazed earthenware; 1 hollow silver button, 1 brass spigot, and 1 red painted clay pipe stem.

It seems obvious from this synthesis of architectural and artifactual evidence that the original fort can be dated to the early 1800s. Several of the architectural features and particularly the bail fasteners can be attributed to the North West Company. Thus, this author feels confident in assigning the original fort to the North West Company strictly on the basis of archaeological data.

The Extended Fort

The second fort at Rocky Mountain House is an enlarged and revamped version of the original fort. However, the changes made are of such a major degree that in fact the impression of a new fort is created. It is also important to note that there is no archaeological evidence indicating anything but a relatively continuous occupation of this site, from the time of the original fort through to the final demolition of the second fort.

Details on the extended version of the forts are copious in comparison to the original structure. The entire southern exterior wall of the original fort was dismantled and extended 16 ft. to the south. This gave the new fort dimensions of 116 ft. north-south by 90 ft. wide, with a much larger area of 10,440 sq. ft. The former eastern, northern and western palisade walls were retained in the new version, as was the north corner bastion. As described in an earlier chapter, the picket trenches dug for the extension walls were fire-reddened as opposed to the uncoloured trenches of the original fort's palisades. This distinctive feature indicates that the extension was constructed during a season of the year when the ground was frozen and had to be thawed.

Included in the plans for enlarging the fort was the erection of a heavy block house bastion constructed of horizontal square beams tenoned into four mortised corner posts. This style of bastion was erected at the south corner of the fort and is distinctively Hudson's Bay Company architecture. The 10 ft. square southwest bastion also has fire-reddened corner post pits, and in one place the north corner pit is intruded by the superimposed trench for the new southern wall of the exterior palisade. This indicates that building of the new bastion commenced prior to the erection of the southern wall of the extension.

The overhead bastion of the original fort was also dismantled and set up over a new gateway in the exterior southern wall of the extended fort. Its large corner support post pits are fire-reddened in both the old and new locations, suggesting that both the dismantling and re-erection activities took place when frost inhibited easy digging of the ground. In effect, this second overhead bastion was a replica of the one which stood over the south entrance of the original fort.

Within the enlarged fort a completely new complex of buildings was erected in the typical architectural style of the Hudson's Bay Company. Buildings IV and V, in particular, overlay area within the original fort and that provided by the extension. Clearly, they date after the extension was made. Buildings II, III and VI are also attributable to the Hudson's Bay Company. It seems probable that they, too, date to the period of enlargement, for they help form a regular and organized building pattern within the fort.

Buildings I and II at the north end of the fort represent habitation quarters, complete with fireplaces. Building II had a large double-hearth fireplace located near the central western end, presumably near sleeping quarters, while an indoor latrine was located at the northeastern end of the building. The west side of the fort appears to have been devoted to buildings used for trading and storing. Building IV probably represents the Indian or trading house from which furs could be taken to the large stationary fur press standing in front of building III. The rough nature of building III and the artifacts found within it suggest that it was a warehouse. Bounding the east side of the fort are the three buildings V, VI and VII. Building V appears to be an earthern-floored workshed with a probable forge at the east end. Building IV has the general appearance of a stable-like structure, and building VII probably represents some type of subterranean cold-storage repository. Whether it was still in use at the time of the extension is not clearly known. Suffice it to say that all of the above buildings encircle an open central 38-ft. square.

In addition to the distinctive Hudson's Bay Company style of architecture, eight silver Hudson's Bay Company buttons were found in the extended fort. Four of these came from within building II and clearly confirm a Hudson's Bay Company occupancy of the fort as well as residence in building II. New types of early stamped nails produced between 1820 and 1830 also appear in the extended fort, although it is obvious that many early types were also utilized. It seems reasonable to believe that nails were salvaged from the dismantled buildings of the original fort and were used in the new extended version. In total, all datable artifacts from the site, particularly the nails, date primarily between 1800 and 1830.

From this data, the author feels very confident in assigning the extended fort to the Hudson's Bay Company, and would date their occupancy between the early 1820s and the 1830s on the basis of the datable nails and other artifacts from the site. The exact date of the enlargement and revamping of the fort is a topic left to the following discussion of historical cross-checks.

Certainly, the extended fort was a much stronger establishment than the original post. The additional bastion at the south corner, the reinforced exterior palisades, the interior line of security pickets, ad a possible watch house outside the fort's southeast corner attest to this observation. There is also ample evidence indicating that the extended fort was completely dismantled and burned at the end of its history. This appears to have been purposeful, and probably offered an opportunity to salvage nails and other pieces of heavy ironware.

Historical Cross-Checks

The historical records contain further data which can be brought to bear on the identification of the two forts at Rocky Mountain House. As noted in the foregoing pages, a North West Company and later Hudson's Bay Company occupation, from the early 1800s to the 1830s, is indicated from the archaeological evidence. Is this consistent with the documentary evidence?

Descriptions of the early version of Rocky Mountain House come from the journals of North West Company employees. David Thompson records that the fort was first established in September, 1799, by the North West Company (Thompson 1962: xlvi), but he does not expressly state who built the fort. This question has posed a problem in past years (McGillivray 1929: App. 1, 6), but Dempsey (1967: 8) has now clarified the matter. John McDonald of Garth is not claiming more than his due when he claims to be the builder of the fort, despite his mistaken memory of the date being 1802.

The journal descriptions of the early Rocky Mountain House are few, but useful information may be gleaned for identification purposes. David Thompson's calculation of the position of the fort at 52°21'20" N. latitude by 114° 58'50" W. longitude (Thompson 1962: liv) is extremely close to the 52° 22'15" N. latitude by 115° 07'00" W. longitude location of this site. Similarly, Alexander Henry the Younger's description of the location of the fort conforms remarkably to the position of the site also. Henry describes the fort on the north bank of the North Saskatchewan River opposite the first major rapids in the river. These general descriptive details are consistent with the location of Rocky Mountain House, but are not strong enough to confirm a positive identification of the fort.

Specific architectural details, however, are sensitive cross-checks for positive identification. Thompson speaks of arranging and erecting an elevated half-bastion with four posts over a gateway in the south end of the fort; this activity began 21 November 1806, and continued through to 3 March 1807 (Dempsey 1967: 49-50); The description of this distinctive type of bastion conforms exactly to the archaeological evidence of the original fort at Rocky Mountain House. This same style of bastion was also retained over the later extended fort's south gateway.

At a later date, on 3 May 1811, Alexander Henry the Younger makes reference to a southeast bastion (Dempsey 1967: 54,67). There is no archaeological evidence to confirm this statement, which leads this author to agree with Dempsey (1967: 67) that if either Thompson or Henry was wrong in their orientations of the south bastion it was probably Henry. Henry does mention later a "Block house over the Gate" (Dempsey 1967: 54), which would be in accord with the south gateway overhead bastion mentioned by Thompson and appearing in the archaeological ground plan of the original fort. Henry also mentions a western gate in the fort (Henry 1897: II, 660), which does not conform to the archaeological evidence.

Frequent reference is made by Henry to multiple bastions at Rocky Mountain House, but he cites no specific number (Henry 1897: II, 642, 658, 666). These structures were "wretched excuses for defense." The archaeological evidence at the site indicates that the original fort had two bastions, one at the north corner and the other over the gateway in the south wall of the external palisade. Multiple bastions, therefore, can be confirmed.

Another specific architectural feature which appears to offer an excellent cross-check is the reference by Thompson to an ice house. Thompson relates that on 16 March 1807, they dug an ice house or glacière in which they placed 90 quarters of meat (Dempsey 1967: 50). This reference again fits the archaeological findings in the original fort. Subterranean building VII in pit 11 in the northeast corner of the fort has been interpreted as a probable cold-storage repository, and the nails from the upper levels of this feature consistently date between 1810 and 1825. This crosscheck is exceedingly convincing.

As noted earlier, archaeological details about the interior of the original fort are meagre. One building and four pits remain the only recognizable features. David Thompson and Alexander Henry the Younger, however, relate other details. There existed a stable (Dempsey 1967: 50); a warehouse (Dempsey 1967: 49); three men's houses (Dempsey 1967: 52); several chimneys (Henry 1897: II, 666); an Indian hall (Dempsey 1967: 53); a forge (Henry 1897: II, 702); a garden in which the inhabitants attempted to grow potatoes (Henry 1897: II, 701); a hen yard (Dempsey 1967: 51), and a flagstaff (Dempsey 1967: 52). Of all of these features, only the factor's house may possibly be correlated with building I of the original fort at Rocky Mountain House.

From the available evidence it is very obvious that close cross-checks exist between the archaeological and historical data with regard to descriptions pertinent to the original fort. There is no doubt that this excavated fort can reasonably be correlated and identified with the North West Company's Rocky Mountain House built inn 1799. The historic descriptions are generally consistent with the archaeological site, and North West Company architecture as well as artifacts were excavated from it. In particular, the overhead bastion over the south gateway and the subterranean cold-storage building of the original fort match similar descriptions in the historic records. The cross-check of these two rather unique and therefore sensitive architectural details serves to corroborate even more convincingly the already established archaeological identification. It is important to note that this important identification rests on concrete evidence and not on conjecture.

There now remains the task of historically identifying the extended and revamped version of the original fort at the site of Rocky Mountain House. The archaeological evidence indicates that the Hudson's Bay Company was responsible for this rebuilding, and did so during the 1820s and 1830s. How does this evidence correlate with the documents?

No historical records document whether Acton House or Rocky Mountain House was the fort retained after the 1821 merger of the North West and Hudson's Bay companies. Dempsey (1967: 15) is of the opinion that Rocky Mountain House was most probably the fort occupied. The archaeological evidence at FcPr-1 substantiates this belief, for there is no indication of anything but a continuous occupation of the fort from the early 1800s into the 1820s. Datable artifacts confirm this statement as does the sequence of extension and rebuilding activities.

Chief trader John Rowand, a former Nor'Wester, took charge of Rocky Mountain House in 1821 and remained until 1823. Apparently the fort was abandoned after this date until Henry Fisher returned to take charge in 1828. From 20 October 1828 to 2 April 1831, constant reference is made in the post journals to repairing and constructing new buildings (Dempsey 1967: 55). Often such references mention "New Houses" and a "New Fort." For instance, on 5 February 1829, "two men began to saw Pickets for a New Fort," and on 12 October 1830, "six men making two chimneys in the new House and four at the flooring." Dempsey (1967: 56) notes that, "It soon becomes apparent that the term 'new' was used in an offhand manner, for seldom did it indicate anything more than the repairing or replacing of part of the old structure."

This is clearly consistent with the archaeological data for the extended version of the fort. As noted earlier, the extension and renovation are of such a major degree that the impression of a new fort is created, in spite of the fact that the extended fort developed as an enlargement of the original post.

There are no pre-1828 descriptions of reconstruction at Rocky Mountain House by the Hudson's Bay Company. Thus it appears reasonable to identify and correlate Henry Fisher's 1828 remodelling activities with the beginning of the extended fort. This identification is consistent with the sequence of architecture, the datable artifacts and the obvious Hudson's Bay Company occupation of the extended fort.

Specific structures of the extended fort can be correlated with those mentioned in the 1828-31 post journals. The sawing of pickets for the "New Fort" and the succeeding construction from 5 February to 20 April 1829 probably refers to the erection of the extension palisades. This activity took place at a time of year when frost is generally found in the ground, and it will be remembered that the trenches dug for the extension walls were fire-reddened. This feature is believed to be a result of thawing the ground with fire, and thus there is consistency between the records and the archaeological findings.

Reference is also made in the records for 12 October 1830 to two chimneys and flooring being laid in the "New House." Building II in the extended fort conforms to this description with its flooring and large double-hearth fireplace. Similarly, building IV offers a probable candidate for the Indian house begun on 15 January 1829. Three bastions in the extended fort offer a number of alternatives for the historical references to these structures, and the reference of 1 March 1830 to "The smith making nails" implies the presence of a forge. Such a feature is inferred for pit 9 in building V within the extended fort. It is interesting to speculate whether the iron cache in pit 14 is the same one mentioned for 29 April 1829, when a hole was dug "in the Bafonts de La Jeuness [literally in the floor of the new one — a building] to Put the Iron work en cache" (Hudson's Bay Company: n.d.). The one distinctive feature, a watch house erected in April 1831, cannot be identified in the archaeological ground plan. This is because extensive area was not excavated beyond the southeast corner of the extended fort where such a structure would most probably have been located. Perhaps a future test excavation could clarify this important detail.

Some confusion has arisen in the past over the duration of the Hudson's Bay Company fort at this location. Its description is obviously incompatible with the later descriptions of Henry Moberly and with Paul Kane's painting, yet the general description given by Hector (Palliser 1863) does obtain. It may be that Moberly is in fact speaking of an entirely different fort from either the extended Hudson's Bay Company version at FcPr-1 or that at FcPr-2. The artifactual evidence from FcPr-1 also substantiates this belief for there is a significant lack of items dating between 1840 and 1860.

Fortunately Dempsey has found a crucial piece of evidence to resolve this problem and establish the terminal occupation date for the extended fort at FcPr-1. In the Rocky Mountain House journals, John Edward Harriot states that a new fort was begun by the Hudson's Bay Company in January, 1835, "a short distance from the old" (Dempsey 1967: 58, 69). This new fort is the Rocky Mountain House of 1835-61 described by Kane, Moberly, Gladstone and Hector. The available archaeological and historical data, therefore, indicates and confirms the identity of the extended fort at FcPr-1 as being the 1828-34 Hudson's Bay Company's Rocky Mountain House. This is the only interpretation compatible with all Hines of evidence.

In all probability, Henry Moberly and Paul Kane are not mistaken, as initially believed by this author, in their descriptions of Rocky Mountain House. The fort they describe is probably located on the same plain as FcPr-1, a short distance to the south. Dempsey (1967: 69) is also of this opinion. During the 1963 excavations Mr. Bill Brierly was disc-harrowing the field south of FcPr-1 and recovered a Hudson's Bay Company iron axe head which he donated. This was found about 1,000 ft. southwest of FcPr-1 near the second telephone pole located at the base of the second terrace on the plain. Possibly this is the location of Rocky Mountain House, 1835-61.

The archaeological and historical identification of the two forts at Rocky Mountain House (FcPr-1) now draws to a close. The original fort can be identified as the North West Company post built and intermittently operated between 1799 and 1821. With the 1821 merger, the Hudson's Bay Company took jurisdiction over the fort and kept it open in the years 1821-23. The fort was then enlarged and rebuilt in 1828-31 under Henry Fisher, and this new fort can be correlated with the extended and revamped version at FcPr-1. Figure 54 represents the ground plans and historical identification of the two forts excavated at Rocky Mountain House.

54 The two forts at Rocky Mountain House. a, the original fort, 1799-1828, North West Company and Hudson's Bay Company; b, the extended fort, 1828-34, Hudson's Bay Company.

With this archaeological and historical evidence compiled, it is now obvious that there are five forts at four sites in the Rocky Mountain House area. These include: the Hudson's Bay Company's Acton House (1799-1821); the North West Company's original Rocky Mountain House (1799-1821) at FcPr-1; the enlarged and revamped Hudson's Bay Company fort at FcPr-1 (1828-34); the later Hudson's Bay Company fort (1835-61), and the final Rocky Mountain House (1866-75) at FcPr-2. This sequence and series of identifications is the direct result of a coordination of archaeological and historical research. The specific identifications made were not possible on the basis of one line of research alone.