Canadian Historic Sites: Occasional Papers in Archaeology and History No. 6
by William C. Noble
Artifact Descriptions (continued)
TRADE AND BUSINESS ITEMS
This category comprises the largest number of artifacts from the site, due entirely to the high frequency of glass trade beads which are counted singly. As a group, the items of trade and business account for 10,861 or 94.7 per cent of the total (11,464) artifact inventory. Glass trade beads alone account for 94.4 per cent. While it is obvious that beads constitute an important trade commodity, there are seven other kinds of commodities also represented. Even so, there are many trade items of a perishable nature that are not preserved at the fort. Such items as coats, caps, blankets, feathers and the like are good examples; these are frequently recorded in trade goods inventories. Table 15 presents an itemization of the trade and business items recovered at the fort.
Glass Trade Beads
Glass trade beads (Fig. 52) have a high incidence on virtually all historic fur trade sites and Rocky Mountain House is no exception. A total of 10,832 glass beads comprise the most numerous artifact commodity excavated from the fort, with 7,094 of these beads coming from a single bead cache. In order to bring some type of standardization into the terminology and classification of the beads from this fort the author has used three reference sources. First is W. C. Orchard's (1929) classic work; the second is G. Hubert Smith's (1953) work from Fort Berthold, North Dakota; and the third is Carl F. Miller's (1960) work on the beads from Fort Lookout, South Dakota. None of these works is entirely consistent with the others, but all contain pertinent data for analysis of beads from western forts.
At the outset, the beads from Rocky Mountain House can be separated according to a size criterion. In keeping with Miller (1960: 63), a "seed bead" is defined as being a bead which measures less than 2 mm. in diameter. Within the Rocky Mountain House sample only 151 beads can be classified as belonging to this type. The diameter for all bead varieties considered represents the width of the bead taken at right angles to its stringing hole.
Notably the seed beads are restricted to the three colours red, white and blue. The 10,684 glass beads larger than 2 mm. come in seven colours. These two categories of beads are compared in the foregoing table together with the bead counts. Clearly, the tallies of the bead colours indicate that white and blue beads are dominant. The inference suggested is that these two colours were preferred by the aboriginal populations. Black and yellow glass beads have the lowest frequencies.
Shapes of glass trade beads are also a common attribute used for classification and a variety of these exists. Four major forms have been recognized and selected for the Rocky Mountain House sample; subcylindrical, tubular, barrel and globular. Within these major forms are ten subvarieties, each designated by a small alphabetical letter. Figure 52 illustrates these forms.
Analysis of the seed beads indicates that all are of subcylindrical shapes a and b. No separation was made to make frequency counts of these two shapes. A perusal of the distributions of the seed beads, however, shows a marked concentration in building I and pit 10. Building I alone produced 104 specimens recovered loose among the floorboards. The following table records the seed bead distributions.
The comparatively low number of seed beads at Rocky Mountain House is in contradistinction to bead samples from other historic forts. For instance, seed beads are by far the most numerous at Fort Michilimackinac (Maxwell and Binford 1961: 90), and, similarly, many appear at Kipp's Post. But it is important to note that Woolworth and Wood (1960: 281) include beads over 2 mm. diameter under their seed bead heading. Miller (1960: 63) records 225 seed beads in a total of 659 beads from Fort Lookout II. This comparative data suggests that the Rocky Mountain House seed bead sample is very low. This may be a function of time, different suppliers or simply due to a preference on the part of the Indians.
Large Subcylindrical Beads
Large subcylindrical beads (i.e., 2.1 mm. to 4 mm. in diameter) dominate the glass bead sample from Rocky Mountain House. A total of 10,633 specimens are of this variety.
The subcylindrical beads are broken down according to three prevalent shapes correlated with colour. All seven colours, white, blue, red, purple, green, black and yellow, are represented in the large subcylindrical glass bead sample, but only beads of shape b carry the entire range. Shape c subcylindrical beads dominate the sample with 10,539 specimens, but are restricted to the three colours red, white and blue. Shape a specimens are only present in white, blue green and purple colours. In total numbers, both shapes a and b are equally represented with only 47 specimens. Table 18 presents the detail of the analysis.
The distribution of large subcylindrical beads over the site clearly indicates the predominance of specimens from the bead cache east of pit 11. This cache of 7,094 beads considerably skews the sample frequencies in favour of white and blue beads which are the only colours represented in the cache. Another heavy concentration of beads occurred in the dump against the east palisade. Excluding these two areas, there is a consistent correlation between bead concentrations and buildings or pits. Very few beads occur loose in random dispersal over the site.
The other subcylindrical beads coloured red, green, purple, black and yellow have a particularly close association with building I and pit 10. These beads may be early. None appear in building II, but within this building there is another distinct distribution. Blue subcylindrical beads are nearly equally distributed within both east and west ends of building II while the white beads show a much heavier concentration in the eastern end. The meaning of this pattern is not clearly understood.
All subcylindrical beads from pits 8 and 11 have late dating associations. Table 19 lists the distributions of coloured subcylindrical beads from the fort.
Tubular Glass Beads
A total of 28 tubular glass beads comes from the site. These range in size from the one large white specimen measuring 2.2 cm. long by 6 mm. wide to smaller specimens measuring 6 mm. long by 3 mm. wide. All have been cut from glass cane tubes.
Two varieties, shapes d and e, are recognized within this general form (Fig. 52). A compilation of the incidence of these two tubular forms together with bead colour indicates that plain white d tubular beads are the most common. Blue tubular beads have an equal representation between the d and e forms. Five of the seven bead colours appear on tubular beads. It should be noted that the green tubular bead is actually a green striped specimen with four thin, straight green stripes running parallel to one another along the length of the bead. The background colour is white.
Tubular glass beads have a limited distribution at Rocky Mountain House. All specimens come from buildings or pits within the north and northeast ends of the fort. The west end of building II produced the largest number, nine, of which eight are white in colour. Distinctively, the only red and purple tubular beads come from building I. Table 21 lists the distribution of tubular beads.
Fifteen specimens are classified as barrel beads. These have three shapes, f, g and h, and appear in only four colours, purple, black, blue and green. The beads are 5.5 mm. to 9.0 mm. long by 3 mm. to 5.5 mm. in diameter. In table 22 it will be noted that purple is the preferred colour.
Lowest in frequency from the site are five globular beads. These are 6.0 mm. long by 5.2 mm. to 7.0 mm. wide. Only two shapes are represented, j and k. Shape k is a distinctive bead having seven parallel flutes running parallel to the long axis of the bead. Blue and purple are the two colours represented by the globular beads.
All in all, the glass trade beads represent the single most numerous artifact item preserved at Rocky Mountain House. The three dominant colours are red, white and blue, which may indicate the European response to a preference on the part of the Indians trading at the fort.
The 7,094 white and blue subcylindrical beads from the bead cache east of pit 11 came from a circular ring feature suggestive of some type of perishable sac or bag. The location of this bead cache and the nearby dump against the east wall of the exterior palisade also suggests the possibility that a bartering widow was present through the eastern palisade in this vicinity. Together these two bead concentrations, which are the only two of their kind in the fort, account for an astonishing 9,396 beads of the total sample of 10,832.
The distribution of barrel beads over the site shows a predominant association with buildings I and II. Others occur in pit 11 and the surface dump adjoining pit 10. These proveniences date early and late.
Distribution of the globular beads is predominantly restricted to pits.
Unfortunately the dating of the beads from Rocky Mountain House must remain relative. No single bead type can unequivocally be assigned an absolute date. For this reason none of the recovered beads can definitely be attributed to the North West Company. However, there are some beads, particularly the seed varieties and the globular fluted beads, which do show a close association with early features in the fort. Seed beads came predominantly from building I and pit 10 which are considered to be earlier than the Hudson's Bay Company occupation and attributable to the North West Company. This same distribution is noted for the globular fluted beads. These beads and those exhibiting the rarer colours of green, purple, black and yellow probably are early varieties. No large blue china ware beads were found as described by Alexander Henry the Younger (Henry 1897: II, 753) during his stay at the fort in 1811.
Metal Projectile Points
Excavated from the fort are ten metal projectile points (Fig. 53, a-i) cut with a chisel from sheet iron, copper and brass. These were probably cut by the traders at the fort for trade to Indians for use in tipping their arrows. As such, the shapes of the points do not necessarily conform to lithic counterparts made by the Indians, but rather represent the trader's conception of what an arrowpoint would look like and what was the easiest style for him to cut. Uniformity is present in the triangular, stemmed shape of all the points. Eight specimens are of iron, one of copper and one of brass. The first specimen illustrated (Fig. 53, a) has been cut from a serrated saw blade or bread knife. Table 27 gives pertinent measurements.
The distribution of cut metal projectile points shows association with four of the buildings and one pit within the fort. Three specimens occurred loose inside and outside of the fort.
Cut metal projectile points have been recovered from other western forts dating between 1820 and 1863, but not in such numbers as at Rocky Mountain House. A single stemmed triangular point of brass is recorded from Kipp's Post (Woolworth and Wood 1960: 282), and a similar type of steel point is known from Fort Smith II (Smith 1960a: 141). The substantial number of cut metal projectile points at Rocky Mountain House suggests that they were a popular and common trading item. David Thompson is recorded as having traded several hundred iron arrowheads to the Interior Salish during the winter of 1809-10 along with "upwards of twenty guns" (Thompson 1962: 305).
Common on many historic sites are brass banglers or tinkling cones made from sheet brass rolled into tapering cones of various lengths (Fig. 42, a-b). Eight of these trade items were recovered at Rocky Mountain House, two of them nested together in the original packing fashion. Such banglers were commonly fastened by the Indians to the borders of their dresses, shirts and leggings.
Three general sizes of brass banglers are represented at the fort. These measure 4.6 cm. to 4.7 cm. long, 4.3 cm. to 4.4 cm. long, and 2.3 cm. to 2.8 cm. long. All specimens are of rolled sheet brass open at each end.
All but one of the brass banglers came from the southern or the western interior of the fort. This is evident distributional table 29. The prevalent concentration was around building IV and the south gateway, with the two nested cones coming from building III, the building which is believed to have been a store house.
Seven of these distinctive kettle parts were recovered during excavation (Fig. 53, j-m). They represent the fixtures used for attaching a wire bail handle to the sides of a copper or brass pail. Significantly, the North West Company used a different style of bail fastener than the Hudson's Bay Company (Walter Kenyon: personal communication). The trade kettles and pails of this latter company had bail fasteners composed of simple flanged lugs driven through two exterior opposite upper sides of the pail. The wire ends of the bail handle were then bent around the exterior lug projections in small loops. An excellent example of this type of bail fastener is illustrated by Woodward (1948: 3).
In contrast, the bail fasteners on North West Company trade kettles were more complex. They utilized two small flat rectangular plates fastened on either side of the pail top. These plates were cut from short lengths of sheet copper or brass, folded in half and then inserted over the sides of the pail rim. Two large-headed rivets were driven through the base of each outside plate, through the pail wall and finally through the interior plate half, where the rivets were flattened to hold the plate fast to the pail. The upper portion of each plate projected above the kettle rim as high as an inch. The two upper projecting corners of each plate were then folded as small triangular ears toward the exterior of the pail and flattened. Finally, a circular hole for insertion of a wire bail handle was punched through each plate above the kettle rim. These holes were punched from the interior to exterior judging from the flattened ragged flanges encircling the exterior margins of the bail holes. This feature and the exterior folded ears left no inward projections on the North West Company style of bail fasteners which might inhibit smooth nesting of pails during transport.
All seven of the recovered bail fasteners from Rocky Mountain House are of the North West Company style. No Hudson's Bay Company style specimens occurred. One small specimen (Fig. 53, k) displays its original mode of attachment to the rim portion of a copper kettle. Of the seven bail fasteners recovered, four are brass and three are copper. It will be noted from Table 30 that there is a general functional correlation between the size of a bail fastening plate and the size of its bail hole. This is to be expected where smaller fastening plates, and by inference smaller pails, would have a lighter bail handle than larger kettles. Of interest is the consistent gauge (thickness) of the plates, which is 0.5 mm., regardless of plate size. Dr. Walter Kenyon (personal communication) has recently analyzed copper and brass kettles recovered from Fort Albany and the Winnipeg River in Ontario. He has noted a consistent correlation between the size of a kettle and its gauge, with the larger kettles having thinner walls. Undoubtedly, this feature accounts in part for the generally poorer state of preservation of large kettles than of smaller specimens.
The measurement data on bail fasteners from Rocky Mountain House indicates a uniformity in gauge for both brass and copper specimens. This also appears to be the case regardless of inferred differences in vessel size. In fact, there is the suggestion that bail fastener gauges were standardized for all sizes of kettles and were simply cut in mass from a common 0.5 mm. thick sheet of copper or brass.
The distribution of the seven bail fasteners from the fort indicates wide dispersal with some concentration in buildings I, II, IV and V. There is no observable distinction between the distribution of copper ad brass specimens except a negative association of copper specimens with Hudson's Bay Company structures. The only copper specimen from a building comes from between the floorboards of building I, which displays North West Company architectural style.
Aside from architectural features, the seven bail fasteners from Rocky Mountain House are the best evidence to confirm North West Company occupation of this site. However, there remains the possibility that North West Company merchandise was handled by the Hudson's Bay Company. Brass specimen 934 suggests this, for it was found in a Hudson's Bay Company context along the north end of the double-hearth fireplace in building II. Similarly, brass specimens 644 and 483 came from within the southern ends of buildings IV and V, also of Hudson's Bay Company provenience. These latter two specimens, however, are incomplete and may have been cut to salvage rivets (Fig. 53, l-m). It is obvious that metal from bail fasteners was not used to cut metal projectile points for the latter are of a heavier gauge. With the Hudson's Bay Company takeover of the fort in 1821, it seems reasonable to believe that any remaining old North West Company stock would be used. The limited evidence from bail fasteners indicates that some brass kettles of North West Company style did remain and were cut up by later Hudson's Bay Company employees.
Portions of two slate pencils give a partial insight into the business affairs at the fort (Fig. 51, i). No doubt they were used for tallying bales of furs and hides.
The broken tip portion illustrated is from a pencil measuring 4.3 cm. long by 5 mm. thick; its hardness according to Mohs' scale is 3. This tip portion has seven longitudinally faceted sides worn down to a circular point. It comes from the lowest level of pit 10 at 54 in.
The second pencil is represented by a middle section 2.9 cm. long by 4 mm. thick. It too has seven irregular longitudinally faceted sides. It was found in the turf in the vicinity of pit 12. The two slate stylus portions do not fit together and are, therefore, considered to represent portions of two distinct pencils.
Slate pencils such as these are known from Fort Stevenson of a later 1867-83 date (Smith 1960b: 214).
This copper bale tag or seal (Fig. 44, i) is a plain circular disc with a small 2 mm. hole punched through its upper border. The diameter of the disc is 2.7 cm. and its thickness is 1 mm. As mentioned, the bale tag is plain with no recognizable marks on it. It is from the fill around post 9 of building III.
This specimen is made of heavy cast brass (Fig. 50, i). It is incomplete and shows many signs of mutilation. The broken tap is locked in a closed position and the end of the spout has been cut off raggedly. The square nubbin pail holder located above and in front of the spout remains intact as does the main spout barrel. The length of the spigot is 7.4 cm. and its exterior diameter is 1.8 cm. The spout orifice has an interior width of 1.1 cm.
This brass spigot was found in the top 6 in. of the turf within the northeasterly corner of building III. It is obviously a portable item for fixture into liquor kegs or other such barrels.