Fort Vancouver Historic Structures Report

History and dimensions

As determined by excavations conducted during the fall of 1947 by Mr. Louis R. Caywood, a National Park Service archeologist, the Fort Vancouver stockade at the time of its greatest extent formed a quadrangle approximately 732 feet long and 325 feet wide. The exact lengths of the walls, as revealed by actual measurements of their remains, were as follows:

Subsequent excavations during the years 1948, 1950, and 1952 revealed that these were not the only palisade walls to surround Fort Vancouver. Inside these exterior limits the remains of other stockade walls were uncovered, indisputably proving that at various periods during the fort's history the size of the area enclosed within the pickets had changed. [2]

The dating and sequence of the several walls have been considered in detail, with a presentation of the related historical and archeological evidence, in an earlier study. [3] There seems to be no need to repeat this material here, but the discovery of additional data during the intervening years makes it possible to be more positive concerning several points.

In the following discussion, therefore, it should be understood that statements unsupported by footnote citations are based on material in Hussey, The History of Fort Vancouver, pp. 118-127, where documentation will be found. Passages in the present text based upon other sources can be identified by the fact that they are footnoted.

In order to follow the discussion presented below it will be necessary for the reader to refer frequently to the plan, "Summary Sheet, Archeological Excavations, Fort Vancouver National Monument," dated September 1, 1954, which is plate I in the present report. Stockade walls designated by letters such as HD and CF will be found delineated on that map.

As nearly as can be determined from available data, the sequence of stockade wall construction at the 1829-1860 site of Fort Vancouver was as follows:

1. Original fort enclosure, 1829. When the post was moved from the bluff down onto the riverside plain during the winter of 1828-1829, the stockade erected at that time enclosed a parallelogram measuring 318 feet north and south and about 320 feet east and west. This is the nearly square enclosure outlined by the letters ABED on the "Summary Sheet, Archeological Excavations, Fort Vancouver National Monument."

Hitherto there has been some question as to whether this square or one of nearly the same size lying directly to the east (BCFE) was constructed first. [4] However, positive proof that the manager's residence of 1841, located in the eastern square, was an entirely different structure from the manager's residence of 1836, though practically identical in appearance, makes it possible to assume that the 1836 structure could have been in the western square and thus removes the chief stumbling block in the way of assigning construction priority to the enclosure ABED. [5]

2. "Doubled-in-size" fort, 1834-1836. Descriptions of Fort Vancouver by at least three visitors between September, 1834, and the fall of 1836 seem to indicate that by the latter date, and possibly by the former, the stockade had been enlarged to about twice its original size. Historical evidence shows that several buildings described as being in the "new" part of the fort were constructed shortly after 1836. These included the new manager's house, or Big House, completed during the winter of 1837-1838, and the Bachelors' Quarters, completed in the fall of 1838. [6]

These structures are indicated on the ground plan of Fort Vancouver made by Lieutenant George Foster Emmons when he visited the post as a member of the United States Exploring Expedition in 1841 (see plate III). [7] They are situated to the east of the original stockade, ABED, and they are within a 318-foot square of palisade walls revealed by archeological excavations and identified as BCFE on the archeological summary sheet. Furthermore, visitors to Fort Vancouver during the summer and fall of 1839 describe the post as being comprised of about 36 buildings grouped to form two courts within the stockade walls. In other words, the interior of the fort was divided by buildings and not by a transverse wall.

It is clear, then, that sometime between 1834 and mid-1839, and almost certainly by 1836, the original 318-foot-square fort was enlarged by adding another square of the same size to it on the east and removing the old wall (BE) between them. The resulting "doubled-in-size" fort (ACFD) measured about 638 feet by 318 feet.

Although the 1841 Emmons map is inaccurate in some particulars -- Emmons evidently did not feel free to make actual measurements -- a comparison of his drawing with the results of archeological excavations indicates that his plan represents the palisade as it stood in the "doubled-in-size" period before any additions were made at the west and east ends.

3. First expansion to the west, 1841-1844. As seems quite evident from the wall locations as given on the Emmons map of 1841 and, particularly, from the relationship of the fort's west wall at that time to the powder magazine and storehouses shown on that drawing, the west wall in 1841 was that designated as AD by the archeologists who found its remains.

The next west wall that can definitely be dated was the outermost west stockade uncovered by the archeologists. This wall, designated as IJ on the archeological summary sheet, was constructed during January and February, 1845. [8] It shows on the "Plan of Fort Vancouver" drawn by Lieutenant Mervin Vavasour of the Royal Engineers during the fall of 1845 (see plate VII) and can be positively identified because it was tied into the blockhouse, the construction date of which was likewise in February, 1845. [9]

However, archeologists in 1952 discovered the remains of a third west wall, designated as HG, lying between inner wall AD of 1829 and outer wall IJ of 1845. Wall HG ran parallel to and about 16 to 18 feet inside of the outer west wall and about 21 feet west of the innermost west wall (AD). [10]

It would seem logical to conclude that this center west wall came between the inner and outer walls in time as it did in space. Yet if inner wall AD existed as late as 1841, as seems almost certain, and if outer wall IJ was built early in 1845, as is demonstrated, then the fort managers must have gone to the expense of constructing the new wall HG and then removing it within the short span of about 3-1/2 years.

A most valuable map in the archives of the Hudson's Bay Company perhaps holds the clue to a more precise dating of wall HG. Entitled "Sketch of Fort Vancouver and Plain, Representing the Line of Fire in September 1844," it shows the fort structures as they existed a year before the Vavasour plan was drawn (see plate V). [11]

Unfortunately, the small scale of this map does not permit one to tell by measurement whether the west wall shown thereon was wall AD or wall HG (it was not wall IJ, since the bastion had not been built by September, 1844). The key is a building shown near the northwest corner of the stockade. If this structure was the Root House which is known to have stood on the same site at a later date, then one can safely say that the west stockade in 1844 was wall HG (see plate I). The Root House, as archeological evidence shows, was built after wall AD had been demolished. But if this structure was the small building (No. 17) identified as a warehouse on the Emmons plan of 1841, then the 1844 west palisade was wall AD. [12] The Vavasour map of 1845 shows no building inside the northwest corner, while the Covington map of 1846 (see plate XIII) shows a small square structure which cannot be positively identified. Since these maps thus give no clear support to the theory that the Root House may have existed as early as September, 1844, it seems impossible to say positively whether the 1844 west wall was AD or HG.

Since it seems extremely unlikely, however, that a new west wall would have been built after September, 1844, and then replaced by a still more westerly wall in January and February, 1845, it seems reasonable to assume that wall HG had already been built when the Line of Fire Map was drawn. If this theory is correct, wall HG was erected between 1841 and September, 1844.

When wall HG was built, the northern and southern palisades were extended about 21 feet westward, creating walls HA and GD and enclosing the space added to the fort by wall HG.

4. Expansion to the east, c. 1844. On the Emmons map of 1841 the east stockade wall is shown as being only a short distance -- about 20 feet-- east of the Bachelors' Quarters. The line of Fire Map, showing conditions as they existed at the very end of September, 1844, places the east wall about 75 or 80 feet east of the same building. This new location corresponds almost exactly with that of the outer eastern wall as revealed by archeological excavations (wall KL less a short addition made when the southern fort wall was later pushed outward about five feet). This shortened version of line KL was found to be, on the average, 56 feet east of wall CF.

It is clear from the Line of Fire Map, then, that the outer east wall was erected prior to September 24, 1844 (the date of the breaking out of the fire). But the construction date can be assigned to a still earlier time. It will be noted that the Line of Fire Map shows a building incorporated in the northern end of the east wall. This structure, as proved by the Vavasour map of 1845 (see plate VII) and other evidence, was the bakery.

The bakery was not completed until after October 15, 1844, but work on it had been under way for some months evidently. On September 18 the fort received a shipment of 5,000 bricks which undoubtedly were for the bake ovens in this structure. [13] It seems most probable that the new east wall was built at about the time the bakery was started, since they apparently were intended to form a unit. [14] Lacking positive evidence, a reasonable guess for the date of the outer east wall's construction would be the spring of 1844. [15] In any event, it can be stated with assurance that the new east wall was built between July 25, 1841, the date of the Emmons map, and September 24, 1844, the approximate date of the developments illustrated by the Line of Fire Map.

Of course when wall CF was moved an average distance of 56 feet to the east, the additional fort area thus created was enclosed on the south by an eastward extension of line EF. The northern end of this space was closed by extending the north stockade by the line CK. This latter wall was a single row of pickets which formed part of the north palisade as it stood at the time of the fort's greatest extent.

5. Second expansion to the west, January-February, 1845. When the Vavasour map of late 1845 is examined closely, it is seen that the north, east, and south walls remained in the same positions as shown on the 1844 Line of Fire Map but that the west wall had been moved outward during the intervening year. As shown by the scale on his map, Vavasour placed this new west wall about 45 to 50 feet west of the Company's trading shop and store. This location corresponds almost exactly with the position of the outer west wall as uncovered by National Park Service archeologists (wall IJ except for the southern four feet).

Confirmation of the hypothesis that Vavasour's west wall was also the extreme west wall of the fort at the time of its greatest extent is found in the fact that Vavasour shows a bastion at the northwest corner of the stockade, exactly where the foundations of a bastion were uncovered in 1947. As far as is known, there was only one bastion in the northwest angle of Fort Vancouver between 1845 and 1860, and no evidence has yet come to light to demonstrate that it was ever moved from the spot upon which it was originally constructed.

Since, as we have seen, the previous west wall, HG, was a relatively new structure, certainly built after July, 1841, it would seem that a new palisade only about 18 feet to the west would not have been erected except in connection with the building of an important structure whose location was determined by factors other than economy. Such a structure was the blockhouse. Evidently it was felt necessary to keep this bastion and a new palisade to be built in connection with it a greater distance from the Company's principal western warehouses and shops. [16]

At any rate, the historical record shows that the construction of the new west wall and the building of the bastion were related in time. Clerk Thomas Lowe noted in his journal on February 7, 1845:

Finished erecting the new Pickets on the West side and part of the North side of the Fort, at which the men have been employed for some time past. A bastion is to be built in the N. W. Corner of the Fort.... [17]

These words seem to fix the date of construction of the outermost west wall, the shortened IJ, with precision. At about the same time the north and south walls must have been extended westward to meet the new palisade. [18]

6. Expansion to the south, 1846-1854. The last change in the dimensions of the stockade occurred when the south wall was moved outward about four to six feet (the distance being a little greater on the east end than on the west) to the position JL. That this move occurred after the completion of the outermost west wall on February 7, 1845, is demonstrated by construction details uncovered during the archeological explorations. For instance the extension of wall DG to the westward would not have been undertaken if there had been in existence an outer south wall which could have been lengthened to close the southern end of the additional fort area created when the outer west wall was built.

There are several known periods of south palisade repair or reconstruction during which this move could have been made. The journal of Thomas Lowe contains the following references to the south, or front, wall:

November 21, 1845. "Some men set to dress pickets for the front of the Fort."

January 20, 1846. "Putting up new Pickets in front of the Fort."

January 23, 1846. "People busy erecting the Pickets in front of the Fort."

February 2, 1848. "Men began putting up new pickets in front of the fort."

January 9, 1850. "Blowing very hard. . . . Last night the wind blew down about 40 feet of the pickets in front of the Fort."

February 18, 1850. "Men employed. . . putting up a row of pickets in front of the Fort, which had fallen down about a month ago." [19]

About a decade and a half after this last entry was written an early settler in the Vancouver region, Lewis Love, testified that between 1850 and 1854 the stockade as a whole was "about rotted down," and that repairs were made. [20]

From these data it seems that after 1845 there apparently were substantial reconstructions of the south wall in January, 1846; February, 1848; and at least once between February, 1850, and 1854. A certain amount of additional information is available which throws light on the possible extent of each of these reconstructions.

a. January, 1846. As is demonstrated in the following section of this chapter dealing with stockade construction details, sometime between the end of 1845 and May 3, 1847, there apparently was a change in the building method used on at least a part of the south wall. Instead of the pickets being fastened to two horizontal girths as had been the case from at least 1841 to the end of 1845, the posts were pegged to only a single girth. Although the evidence of this change is only known to apply to a very short segment of the wall (see plate XII), it, together with Lowe's journal entries, seems to indicate that the January, 1846, reconstruction was a major effort.

b. February, 1848. As is shown by Lowe's journal, the erection of "new pickets" was commenced "in front of the fort" on February 2, 1848. But only about a week earlier, on January 24, 1848, the same industrious clerk had recorded, "A Bastion has been put up to day in front of the Fort." [21]

Thus far no conclusive evidence has been found to indicate the exact location of this blockhouse of 1848, the second such structure to be erected at Fort Vancouver. During the 1952 archeological excavations the remains of "three parallel timbers roughly 6 to 8 inches square" were found where the easternmost 17 feet of the inner south wall may have once stood (see plate I). Mr. Caywood believed that these timbers marked the foundation of the second blockhouse and fixed its location as the southeast corner. [22]

As shall be seen in a later chapter, there is historical evidence tending to confirm this hypothesis, but in the opinion of the present writer the location of the bastion erected in 1848 is still uncertain. Further archeological excavations might settle this question.

In any event, even if the three timbers should be shown to be part of the bastion, they do not shed much light on the problem of whether the south wall was moved outward about six feet when the new blockhouse was constructed. Apparently a foundation in the inner wall location would not have been incompatible with a palisade at either the inner or outer wall sites. If precedent at the first bastion was followed at the second, however, it seems more probable that the inner bastion wall would have been inside the stockade line. It is quite possible, then, that the stockade construction in February, 1848, was designed to advance the south wall to position JL.

c. 1850-1854. It seems quite evident from Lowe's journal that the work performed on the south wall in February, 1850, was more in the nature of repair than complete reconstruction. However, there seem to be grounds for believing that by 1854 the outward movement of the front palisade had been accomplished.

Unfortunately, the small scales of the available maps and the differences between the several copies of them make it impossible to detect a change in stockade dimensions as small as six feet, at least with any certainty. All one can say is that when the Vavasour plan of 1845 (see plate VII) is compared with the careful survey made by Lt. Col. B. L. E. Bonneville in 1854 (plate XIX) the south wall seems in the latter to be farther away from the storehouses inside the south palisade. Almost certainly the Bonneville map represents the fort as it stood at the time of its greatest extent. [23]

One circumstance which seems to support, though by no means prove, the hypothesis that the movement of the south wall had been completed by 1854 is the fact that the location of one of the gates in that palisade was shifted sometime between 1846 and 1854. Of course such a shift need not necessarily have been associated with a movement of the entire stockade wall, but if a change in a gate had been contemplated, it probably would have been easier to make the shift at a time when the entire wall was being reconstructed.

An examination of the Warre plan of 1845 (plate VII) and the Covington map of 1846 (plate XIII) will show that the east gate in the south wall at these dates was directly or almost directly south of the north wall gate. On the Bonneville map of 1854, however, the east gate in the south wall has shifted to the west a substantial distance. [24] This same shift is shown on maps of 1859 and 1860. [25]

The only thing this chain of events proves is that between late 1846, when the Covington map seems to have been drawn, and 1854 there was a change in the location of the southeast gate. But if there is any validity to the theory that the shift in gate location was associated with the outward movement of the south wall, then the latter event can be placed between late 1846 (the Covington map does not show the old Catholic Church which was demolished during June of that year) and 1854.

A review of what is known about the moving of the front wall six feet to the south leads to the following conclusives:

a. The inner wall (line GF as extended to the outer palisades at each end) was the outer south stockade wall when the outermost west wall was completed on February 7, 1845.

b. The rebuilding of the south wall during January 1846, could have involved moving that palisade six feet southward but probably did not. The latter surmise is based on the fact that the Covington map of late 1846 continues to show the southeast gate in the same position as does the Vavasour map of 1845. Also, the south wall, as far as can be determined from general appearance. seems in 1846 to be as close to the buildings inside the wall as it was on the 1845 map.

c. The stockade construction in February, 1848, appears to have been linked with the erection of a bastion somewhere along the south extremity of the fort. It seems quite probable that the south wall was moved to position JL in connection with the construction of the new blockhouse. Whether the southeast gate was moved westward at the same time is not apparent from available evidence.

d. Between early 1850 and 1854 the south wall underwent, at the very least, extensive repair. This activity could have involved the moving of both the stockade and the southeast gate. A reason for assigning the outward movement of the wall to this period might be the fact (which shall be brought out in a later chapter) that the second bastion was a very short-lived structure. Built in 1848, it seems to have disappeared at least by 1854. The removal could have occurred as the result of a rebuilding of the south wall over its site.

On the basis of these conclusions, it appears to the present writer that the inner south palisade most likely formed the south stockade wall in late 1845 and early 1846, the period to which it is intended to restore Fort Vancouver.

Construction details

The stockade which enclosed and protected the major structures at Fort Vancouver was formed of logs which were ranged vertically as pickets or pales. Archeological excavations have confirmed conclusions that can be drawn from the natural distribution of forest trees and from historical evidence to the effect that the palisade logs were all or nearly all Douglas fir. [26]

According to an employee who resided at Fort Vancouver for a number of years only "very choice" logs were used for pickets. When the palisade was first constructed in 1829 there probably was a sufficiency of suitable timber within a reasonable distance of the building site; but in later years, when rotted timbers were replaced or when new walls were built during the periodic fort enlargements, it was necessary to go "a great distance from the fort" to obtain satisfactory timber. The logs were cut, dragged by oxen to the Columbia, rafted downstream, and then hauled again by oxen to the post. [27]

Three visitors who were at Fort Vancouver in 1836, 1841, and respectively, described the pickets as being about eight or ten inches in diameter. [28] Ends of posts found in the ground during the 1947 excavations measured between five and thirteen inches, with the larger posts being situated at the stockade corners. [29] In 1966 careful archeological salvage work was conducted along the outermost north palisade wall, which probably was constructed during January and February, 1845. The pickets in this line ranged from five to ten inches in diameter, the average being 7.2 inches. There were about 120 posts in every 100 feet of wall. No evidence of bark was found but gaps between palisade butts averaged 2.8 inches, a fact which led archeologists to speculate that the posts may have been installed "unskinned" and that the bark may have decayed quite rapidly and completely. [30]

On the other hand, there seems to be no historical evidence concerning whether the pickets at Fort Vancouver were installed with the bark on or with it peeled off. Existing photographs and drawings of the establishment are not sufficiently clear to throw light upon this matter, nor do pictures of the original stockades at other Pacific Coast forts of the Hudson's Bay Company permit a positive conclusion as to the general practice in this regard (see plates XXXI, XXXII, XXXIII, XXXIV). On the whole, however, the pictorial evidence seems to show peeled pickets more often than unpeeled.

If events at Fort Chimo, in Ungava on the eastern side of Hudson Bay, are true indicators, the Company's employees on the East Coast ordinarily peeled the pickets before installing them. During the spring of 1832 the men at that recently established post spent several weeks "peeling the bark off piquets" prior to setting them in place. After the skinning was well under way, entries in the fort manager's journal indicate that the carpenter was set to work "pointing piquets." [31]

Thus, on the basis of the pictorial evidence and of the practice elsewhere, it would not have been uncharacteristic for the pickets at Fort Vancouver to have been peeled.

The length of the posts appears to have varied according to the date at which they were cut. Visitors to the depot prior to the winter of 1844-1845 generally give the height of the stockade as between 20 and 25 feet, although Captain Edward Belcher of the Royal Navy, who visited the fort in August, 1839, stated that the pickets were 18 feet high, "composed of roughly split pine logs. [32]

Those describing the palisade in 1845 and later give figures which range from 12 to 20 feet, with 15 feet as the most frequent estimate. Lieutenant Mervin Vavasour of the Royal Engineers, who made a rather careful plan of Fort Vancouver late in that year, specifically stated that the pickets were 15 feet high. [33] Since Vavasour was a trained observer carefully assessing defensive features, his figures must be accepted unless more convincing evidence to the contrary is revealed at some future date. It seems evident that when much of the stockade was renewed during the winter of 1844-1845, the posts were not cut as long as they had been previously.

Seemingly this decrease in the height of the walls continued progressively during the years between 1845 and 1860. A photograph taken during the latter year apparently shows that the stockade was only eight or nine feet high in places, although at least part of the west wall seems to have been ten or twelve feet in height (see plate XXXIV).

Much of this decrease was due to the method of repair employed, particularly during the years of declining economic activity between 1850 and 1860. A stockade post ordinarily lasted for about four or five years. By the end of that period it would be so rotted at the surface of the ground that it would have to be replaced. As a consequence, new pickets were inserted in the walls nearly every year. [34] But occasionally such repairs were neglected for considerable lengths of time. It was said, for instance, that between 1850 and 1854 the palisade was "about rotted down." Repairs were made during that period simply by cutting off the existing posts and resetting them in the ground. [35]

In addition to the length of the logs exposed above ground, several feet were buried in the earth. It was the usual custom at Hudson's Bay posts west of the Rockies to plant pickets about four feet in the ground, and several visitors to Fort Vancouver say this same procedure was followed at that establishment. [36] But Lieutenant Emmons in 1841 noted that the posts were buried only two or three feet in the ground. [37] Evidently Emmons was a more accurate observer than the other witnesses, for archeological excavations confirm his report. Mr. Caywood in 1947 found the posts planted to a depth of between two and three feet, exactly as reported by Emmons. [38] Mr. John D. Combes, who dug along the north wall in 1966, reported that the posts extended from 2-1/2 to 3 feet below the historic ground surface. [39]

According to at least one witness and in accordance with what one long-time employee considered the usual Company practice on the West Coast, the logs, after being cut to size, were prepared for use as pickets by being sharpened to a point at one end. [40] And, indeed, pickets with pointed tops were used at forts on the Pacific Slope. They clearly show in an 1860's drawing of Fort St. James (see plate XXXV) and in an early photograph showing the outer stockade of Fort Rupert (see plate XXXVI). Further, one of the best-known and seemingly most accurate views of Fort Vancouver, the lithograph by Henry J. Warre showing the establishment as it appeared in 1845, depicts the palisade posts as being conspicuously and fiercely pointed (see plate IX). [41] Ordinarily such a drawing by an eyewitness would be considered conclusive.

Yet, it must be admitted that an impressive case can be made for the thesis that the palisade posts at Fort Vancouver in 1845 were not pointed but were cut off flat or with a slight tilt toward one side, most probably toward the inner edge. In the first place, the use of pointed pickets was by no means a standard practice at Company posts on the West Coast or, for that matter, across North America. Photographs of the stockades at Fort Langley and at Fort Victoria clearly show that the posts were flat on top, at least during the 1860's (see plates XXXI and XXXII). [42] The main stockade at Fort Rupert was not only level on top but was protected by a cap of horizontal timber or logs (see plate XXXVI). [43] According to widely held but undocumented theory, pointed posts fell into disfavor because pilferers or hostile natives could easily loop ropes over them and thus scale the walls.

In the second place, there is specific evidence that the pickets at Fort Vancouver were not pointed at least at certain periods. Lieutenant George Foster Emmons of the United States Exploring Expedition made e careful examination of the fort walls during July, 1841. The following sketch which he made on the spot clearly shows that the pickets were cut off slightly on the bias, with the flat tops sloping slightly towards the inside of the fort.

Figure 1.

(From Emmons, Journal, MS, III, entry for July 25, 1841. Courtesy of the Beinecke Rare Book and Manuscript Library, Yale University.)

The next clear view of flat-topped palisade posts at Fort Vancouver is provided by a photograph taken by the British Boundary Commission party in 1860 (see plate XXXIV). No known view of the establishment between 1841 and 1860, except the Warre lithograph already mentioned, is sufficiently detailed to throw light upon the subject. But in view of the periodic rebuildings of the stockade, one would be rash to assert that because the posts of 1841 and those of 1860 were both flat-topped, there were no periods in between when the pickets were pointed, particularly in view of the Warre drawings. [44] On the other hand, it would seem reasonable that a practice once started would be continued, particularly since cutting the posts across with a saw undoubtedly was easier and less time-consuming than pointing the ends.

The crucial evidence as to the situation in 1845, then, would appear to be the Warre lithograph with its clearly defined pointed pickets. The original pencil sketch upon which the lithograph clearly was based seems also to show pointed posts, although the representation is not as precise as one would wish.

However, the water color which evidently was prepared by Warre as e guide for the lithographer is preserved in the Public Archives of Canada; it shows not the slightest sign of pointed palisade posts. Rather, the top of the pickets is shown as a perfectly level line (see plate X). Therefore, there is the possibility that the pointed pickets were among a number of changes and "improvements" made during the engraving process. These alterations may have been guided by Warre personally or they may simply have represented the engraver's idea of how a proper fur-trading post should look.

On the basis of present knowledge it is impossible to say positively whether the Fort Vancouver palisades were pointed or unpointed in 1845.

If general Hudson's Bay Company practice was followed at Fort Vancouver, the tops of the posts alternately came on the thin and thick ends, so that when placed side by side in the palisade with the pointed or slanted ends up, the pickets would fit together without large gaps, as would have been the case if all the thin ends had been placed up or down. At many Company posts it was ordinary procedure to square two sides of each log so that the pickets would butt together tightly. [45]

There were other methods of assuring that there would not be gaps between the posts. In 1834, shortly after the palisade at Fort Nisqually was completed, it was recorded in the post journal that "Betwixt each Picket of the Fort small poles were put in order to stop the Indians from looking inside." [46]

It is not known whether any of these practices were followed at Fort Vancouver. On November 21, 1845, Clerk Thomas Lowe noted in his diary: "Some men set to dress pickets for the front of the fort." [47] These words clearly indicate that the logs were given some sort of processing before being set in place, but it would be rash to draw from them any conclusions as to the type of treatment given the pickets.

According to evidence uncovered during the 1947 excavations, the ends of the stockade posts which were buried in the ground were saw-cut and were not sharpened. [48] If the Company's lawyers were correct in statements made in the 1860's, however, the buried ends were not put into the ground without any preparation. The usual practice, as intimated by their cross-questioning, was to strip the bark from the ends to be planted and to char them thoroughly on the outside. This procedure evidently helped to preserve the posts from rotting. [49]

After the posts were fully prepared, the next step in stockade construction is not entirely clear. According to the reminiscences of one old employee, the usual Company practice was to attach the pickets to cross pieces or girths which ran horizontally around the inside of the wall about four feet from the top. The posts were fastened to this girth by wooden pegs or by means of an "oblique notch", as illustrated below:

Figure 2.

(After Compton, Forts and Fort Life in New Caledonia, MS, 6.)

The ends of the cross pieces, which were about 15 feet long, were mortised into larger pickets called "king posts." [50]

According to entries in the Fort Nisqually Journal of Occurrences, as nearly as they can be interpreted, the sequence of events by which this assemblage was obtained was about as follows: after dressing and sawing the pickets, the laborers were engaged in "mortising and laying" them; then they apparently were busy "arranging the pickets on the ground," in "arranging the pickets," and in "arranging & boring the pickets." When the "arranging and boring of the pickets" was completed, the next entries record the digging of the "trench in which the pickets are to be placed" and then the "erecting of the pickets." [51]

These entries unfortunately do not answer all the questions that could be raised. For instance, were the mortised, arranged, and bored pickets erected individually or were they first fastened to the girths and raised in sections? Because of the great weight of the green logs, the former procedure probably was employed. Present-day architects believe the construction sequence was about as follows: posts notched for girths and bored for pegs; king posts erected and connected by girths; posts raised to position in the fort wall and notches fitted to girths; posts trenailed to the girths.

Evidently it was common practice to insert wedges in the ends of the wooden pins by which the pickets were fastened to the girths. At Fort Nisqually in 1849, for instance, the post journal noted that on May 14 two laborers named Keva and Kalama were busy "wedging and sawing off ends of picket pins." [52]

That this general type of wall construction -- one line of girths about two to four feet below the top of the pickets -- was widespread at Hudson's Bay Company forts west of the Rocky Mountains is amply demonstrated by drawings and photographs of the stockades at a number of establishments. Pictures of Fort Victoria, Fort Langley, and Fort Rupert clearly illustrate this point (see plates XXXIII, XXXVI, and XXXVII) [53]

This same general type of construction was followed at Fort Vancouver, but with certain important variations. When Lieutenant Emmons examined the stockade there in 1841, he noted that there were two sets of horizontal girths running around the inside of the palisade, one three to four feet above the ground and the other a foot or two below the tops of the pickets. For additional support, necessary because the posts quickly rotted at ground level, diagonal bracing timbers ran at intervals from the upper girth to the ground. [54]

Joel Palmer, an emigrant from the United States who reached Oregon very late in 1845 and left the next year to return east, found conditions much the same during his brief visit. Describing the stockade at Fort Vancouver he said: "A notch is cut out of each log near the top and bottom, into which a girth is fitted, and mortised into a large log at each end, the whole being trenailed to this girth." [55]

In view of this evidence, it seems clear that at least as late as November or December, 1845, or early 1846, the stockade at Fort Vancouver, or a significant part of it, was characterized by a double set of girths as shown in Emmons's 1841 diagram. But very shortly thereafter the construction of at least part of the stockade had changed to the more usual Hudson's Bay type. A water color sketch of the interior of the Vancouver stockade found in the London archives of the Hudson's Bay Company distinctly shows a small portion of the south palisade. There is only one girth, and that is located near the top of the wall, evidently about two feet or less below the tops of the posts (see plate XII).

It has generally been though that this picture might have been drawn in the early 1850's. [56] It seems very probable to the present writer, however, that this picture can be dated between June 18, 1846, and May 3, 1847. [57] If this surmise is true, it seems evident that when the new south or front wall was constructed early in 1846 only one girth was used. [58]

The use of the one-girth construction is confirmed by one of the 1860 photographs of Fort Vancouver (see plate XXXIV). This picture clearly shows that on at least part of the stockade there was but one set of girths and that this line of horizontal cross pieces was four or five feet below the tops of the pickets.

Archeological excavations have thus far not provided clear evidence of the use of king posts or indicated how such posts were spaced. [59] Yet the photograph of 1860 shows that king posts, into which the girths were mortised, were at that time employed at Fort Vancouver. There is some question as to how the king posts were placed since the use of larger posts at regular intervals is not evident on drawings and photographs of the exterior faces of stockade walls at Company posts. This fact brings up the possibility that the king posts may have been set back of the line of palisades (see plates XXXIII, XXXVI, and XXXVIII), though in view of the absence of confirming evidence, such a possibility seems remote.

As far as is known to the present writer, no list of materials used in the construction of a Hudson's Bay Company post stockade on the Pacific Slope is extant. In the fall of 1800 Alexander Henry, of the North West Company, built a fort on the Park River, a tributary of the Red River. Although we know nothing of the appearance of this post, his list of "Wood used in our Establishment at Park River" may be some use to architects working on the proposed restoration of Fort Vancouver. The materials used in constructing Henry's fort walls and bastions were as follows:

Recommendations

a. The south stockade wall of a Fort Vancouver restored to the conditions of late 1845 or early 1846 should be the inner south palisade as uncovered by archeological excavations. At least planning should be done on this basis, leaving the possibility of switching to the outer wall (line JL on the Summary Sheet, Archeological Excavations) should further historical or archeological research reveal that the south wall was moved southward about six feet during the 1846 rebuilding.

b. All other reconstructed walls should be in the outermost locations revealed by archeological excavations in 1947-1952.

c. In an attempt to throw additional light upon the date at which the south wall was moved outward, future archeological excavations should include the following steps:

(1). Excavate the easternmost 17 feet of the inner south wall to see if there is evidence of stockade posts beneath the three parallel timbers found by Mr. Caywood in 1952.

(2). Search for gate locations along the entire length of both south walls.

(3). Search outside the southeast stockade corner for evidence of a blockhouse.

d. An archeological search should be made for the evidence of the king posts, with particular attention to the position of these posts in relation to the smaller pickets.

e. Upon a reassessment of the historical and archeological evidence available in 1966 and on the basis of new information garnered since that time, a stockade restored to the conditions of late 1845 or early 1846 should possess the following characteristics:

(1). Height of posts above ground level: 15 feet.

(2). Logs peeled, with diameters ranging from 5 to 10 inches for ordinary palisade posts and from 10 to 13 inches for king posts.

(3). The tops of the posts in at least one or two walls should be saw-cut, with the flat tops slanting slightly toward the inside of the palisade.

(4). If the date to which the fort is to be reconstructed is prior to January, 1846, the entire stockade should have two lines of girths as shown in the Emmons sketch. If the date is after January, 1846, the south wall should have only one line of girths. This line should be about two feet below the tops of the posts.

1. Louis R. Caywood, Exploratory Excavations at Fort Vancouver, 1947 (mimeographed, San Francisco: U. S. Department of the Interior, National Park Service, [1947], 2.

2. Louis R. Caywood, Final Report, Fort Vancouver Excavations (mimeographed, San Francisco: U. S. Department of the Interior, National Park Service, July 1, 1955), 27-30.

3. John A. Hussey, The History of Fort Vancouver and Its Physical Structure ([Tacoma]: Washington State Historical Society, [1957]), 118-127.

4. Caywood, Final Report, 28; Hussey, History of Fort Vancouver, 144-146.

5. For a presentation of the evidence concerning the new manager's residence see below, Chapter IX, "The Big House."

6. Francis Ermatinger to Edward [Ermatinger], Colvile, March 19, 1838, in Francis Ermatinger, Letters of Francis Ermatinger, 1823-1853, MS, p. [121], in the Henry E. Huntington Library and Art Gallery; James Douglas to Governor and Committee, Fort Vancouver, October 18, 1838, in E. E. Rich, ed., The Letters of John McLoughlin from Fort Vancouver to the Governor and Committee, First Series, 1825-38 (Publications of the Champlain Society, Hudson's Bay Series, IV, Toronto, 1941) (hereafter cited as H. B. S., IV), 260.

7. George Foster Emmons, Journal Kept While Attached to the South Sea Surveying & Exploring Expedition. . ., MS, III, entry for July 25, 1841, in Beinecke Rare Book and Manuscript Library, Yale University.

8. Thomas Lowe, Private Journal Kept at Fort Vancouver, Columbia River [1843-1850], MS, 12, typescript in Provincial Archives of British Columbia, Victoria, B. C.

9. See Chapter III below on Bastion.

10. Caywood, Final Report, 28-30.

11. This 1844 map is henceforth referred to in this report as the "Line of Fire Map."

12. The problem of the succession of structures in the northwest corner of the fort is discussed in greater detail below in Chapter XXV on the Root House.

13. Lowe, Private Journal, MS, 5, 7.

14. Of course, it is entirely possible that the new east wall existed before the bakery was commenced, the latter structure merely being fitted into a gap made in the stockade. No archeological evidence of such a wall was found.

15. It is probable that this construction was prior to June 8, 1844, since the journal of Thomas Lowe, which resumes on that date after a gap from October 1, 1843, does not mention any work on the east palisade. Lowe ordinarily noted such labors.

16. On March 20, 1845, Chief Factor John McLoughlin, in charge of Port Vancouver, wrote to Governor George Simpson: "In the month of January last, some Americans seeing us repair our pickets erect a bastion. . .spread a report. . .we were fortifying the Fort." Robert Carlton Clark, History of the Willamette Valley, Oregon (3 vols., Chicago, 1927), I, 809. The word "repair" may indicate that the existing west wall had already suffered from rot and that McLoughlin may thus not have been hesitant to move it outward to connect with a suitably sited bastion.

17. Lowe, Private Journal, MS, 12.

18. Since Lowe seems to indicate that the bastion had not been started at the time the new northwest corner was enclosed, it must be assumed that some sort of temporary palisade was erected between points H and I. Perhaps further archeological excavations will throw light on this matter.

19. Lowe, Private Journal, MS, 29, 33, 64, 72, 73-74.

20. British and American Joint Commission for the Final Settlement of the Claims of the Hudson's Bay and Puget's Sound Agricultural Companies, [Papers] (14 vols., Washington; Montreal, 1865-1869) (hereafter cited as Br. & Am. Joint Comm., Papers), [VIII], 237.

21. Lowe, Private Journal, MS, 64.

22. Caywood, Final Report, 8-9.

23. Ordinarily one would expect a survey as carefully executed as that by Col. Bonneville to provide the exact dimensions of a structure as prominent as the Fort Vancouver stockade. But such is not the case. C. A. Homan, a civil engineer, later calculated the lengths of the north and west walls as shown by Bonneville to be 724.2 feet and 340.8 feet, respectively. The two measurements as shown by archeological excavations were 731 feet and 326 feet. Caywood, Exploratory Excavations. . .1947, 9.

24. Hussey, History of Fort Vancouver, plate XV (plate XIX of the present report). However, another copy of the same map continues to show the southwest gate as being directly south of the north wall gate. Ibid., plate XVI.

25. Ibid., plates XXI and XXIV (plates XXIV and XXX of the present report).

26. Caywood, Exploratory Excavations at Fort Vancouver, 12. For a more detailed discussion of the evidence on this point, see Hussey, History of Fort Vancouver, 127-128.

27. Testimony of D. Mactavish, in Br. & Am. Joint Comm., Papers, [XI], 71.

28. William Henry Gray, A History of Oregon, 1792-1849, Drawn from Personal Observation and Authentic Information (Portland Oregon; and New York, 1870), 150; Emmons, Journal, MS, III, entry for July 25, 1841; testimony of J. W. Nesmith, in Br. & Am. Joint Comm., Papers, [IX], 35.

29. Caywood, Exploratory Excavations at Fort Vancouver, 12.

30. John D. Combes, A Report of the Fort Vancouver Archeological Excavations of the North Wall (Processed, Vancouver, Washington, 1966), pp. 3-4 and fig. 1.

31. K. G. Davies, ed., Northern Quebec and Labrador Journals and Correspondence, 1819-35 (Publications of the Hudson's Bay Record Society, vol. XXIX, London, 1963), 165-166.

32. Edward Belcher, Narrative of a Voyage Round the World. . . 1836-1842 (2 vols., London, 1843), I, 294. For testimony on this topic by a list of visitors, see sources cited in Hussey, History of Fort Vancouver, p. 129, note 20.

33. Joseph Schafer, ed., "Documents Relative to Warre and Vavasour's Military Reconnoissance in Oregon, 1845-6," in Quarterly of the Oregon Historical Society (hereafter cited as OHQ), X (March, 1909), 46, 85, and plan ff. p. 100.

34. Testimony of J. Nesmith, in Br. & Am. Joint Comm., Papers, [IX], 35; testimony of L. Brooke, in ibid., [VIII], 128.

35. Testimony of L. Love, in ibid., [VIII], 237.

36. P. N. Compton Forts and Fort Life in New Caledonia under Hudson's Bay Company Regime, MS, 6, in the Bancroft Library, University of California, Berkeley; Gray, op. cit. 150; Joel Palmer, Journal of Travels over the Rocky Mountains. . .Made During the Years 1845 and 1846. . . (Reuben Gold Thwaites, ed., Early Western Travels, vol. XXX, Cleveland, Ohio, 1906), 209.

37. Emmons, Journal, MS, III, entry for July 25, 1841.

38. Caywood, Exploratory Excavations at Fort Vancouver, 13.

39. Combes, A Report of. . .Excavations of the North Wall, 4.

40. Testimony of Thomas Nelson, Br. & Am. Joint Comm., Papers, [IX], 88; Compton, Forts and Fort Life, MS, 6.

41. What appears to be Warre's original on-the-spot pencil sketch which was the basis of the lithograph is now preserved in the American Antiquarian Society, Worcester, Massachusetts. It is reproduced as plate 40 in Henry J. Warre, Sketches in North America and the Oregon Territory, By Captain H. Warre, with introduction by Archibald Henna, Jr. (Barre, Massachusetts: Imprint Society, 1970) The pickets seem to be shown as pointed in this drawing, although details in this regard are not as clear as could be desired. It is possible to interpret this picture as showing an uneven row of flat-topped pickets, although more probably pointed pickets are intended.

42. For another Pacific Slope example see the 1860-1861 photograph of Fort Colvile in Erwin N. Thompson, Grand Portage National Monument, Great Hall Historic Structures Report, History Data Section (Multilithed, Washington, D. C.: National Park Service, May, 1970), illustration 22.

43. For an example of this type of construction at a post east of the Rockies see the restored Rocky Mountain House, Heritage Park, Calgary.

44. Thomas Nelson, who visited Fort Vancouver in 1851-1852, later definitely stated that at that time the upper ends of the pickets were "sharpened." Br. & Am. Joint Comm., Papers, [IX], 88.

45. Compton, Forts and Fort Life, MS, 6.

46. Clarence B. Bagley, ed., "Journal of Occurrences at Nisqually House," in Washington Historical Quarterly, VII (April, 1916), 147. Possibly the use of a similar technique by the North West Company at the post built by Alexander Henry on Park River in 1800 is indicated by the fact that the list of materials used included 564 "stockades" 15 feet long and 564 "stockades. . .for rembrits" 8 feet long. Elliott Coues, ed., New Light on the Early History of the Greater Northwest: The Manuscript Journals of Alexander Henry. . . and of David Thompson. . . 1799-11 . . . Reprint ed., 2 vols., Minneapolis, 1965), I, 123.

47. Lowe, Private Journal, MS, 29.

48. Caywood, Exploratory Excavations at Fort Vancouver, 12.

49. Br. & Am. Joint Comm., Papers, [IX], 35.

50. Compton, Forts and Fort Life, MS, 6.

51. Bagley, "Journal of Occurrences," in Washington Historical Quarterly, VI (July, 1915), 192, 193, 194; (October, 1915), 267, 268.

52. Victor J. Farrar, ed., "The Nisqually Journal," in Washington Historical Quarterly, X (July, 1919), 216, 217-218. The purpose of the wedging, evidently, was to keep the pegs tight as they dried out and shrank. It is sometimes said, upon what authority is not known, that square pegs were driven into round holes to make a tight fit and/or to prevent splitting if the pegs swelled upon absorbing moisture.

53. A Committee of assessors noted in 1866 that the pickets at Fort Colvile on the upper Columbia River were "ten feet high and pinned near the top to horizontal timbers." Br. & Am. Joint Comm., Papers, [VIII] 276.

54. Emmons, Journal, MS, III, entry for July 25, 1841. Emmons stated that the girths were placed on either side of the palisade, but in this assertion he was clearly in error. A drawing of Fort Vancouver made by one of his fellow officers, Lieutenant Henry Eld, shows that there were no girths on the exterior face of the stockade (see plate IV), nor are such girths shown on any other known pictures of Fort Vancouver.

55. Palmer, Journal, 209.

56. Glyndwr Williams, "Highlights of the First 200 Years of the Hudson's Bay Company," in The Beaver, Outfit 301 (Autumn, 1970), 52.

57. A note in Hudson's Bay Company, Catalogue of Pictures in Beaver House, London (typescript), in the Hudson's Bay Company Archives states, on the basis of family tradition, that this picture was drawn by Lieutenant Coode, of H. M. S. Modeste. Since this ship sailed from Fort Vancouver for the last time on May 3, 1847, the picture could not have been made after that date if Coode was the artist. Barry M. Gough The Royal Navy and the Northwest Coast of North America, 1810—1914. . . (Vancouver, B. C.: University of British Columbia Press, 1971), 82. The sketch does not show the Old Catholic Church, which was torn down on June 18, 1846. Lowe, Private Journal, MS, 42. Therefore the picture must have been made after that date.

58. Lowe, Private Journal, MS, 33.

59. Combes, A Report of . . . Excavations of the North Wall, figure 4.

60. Coues, New Light on the Early History of the Greater Northwest, I, 123.