Artifacts B.C. Artifacts B.C. - Kosapsom


Table of

Table of

& History






Soil Profile

Time Periods


Uvic Report
at DcRu4






University of Victoria Report
on the Kosapsom Archaeological Dig, 1995

This report concerns the second season of excavation at DcRu4, a large shell midden site in the Municipality of Saanich's Kosapsom Park and in the grounds of the Province of British Columbia's Heritage Properties Branch's Old Craigflower School. The portion of the site examined is situated on the Gorge waterway at Admiral's Bridge, Southeast of the junction of Admiral's and Gorge Roads. Excavations were, again, conducted as part of a University of Victoria archaeological field school with continuing involvement of the Archaeological Society of B.C. The field project ran from May 9th to June 23, 1995.

Twenty-three 1 x 1 m excavation units were opened in the preceding season, randomly selected from grid squares falling on a meandering 3m wide strip that approximated the route of a proposed extension to the Gorge walkway. Excavations in 1995 were similarly confined to this strip. The 1994 excavations suggested that the strip contained materials relating to at least three periods of occupation and use. These were associated with the Locarno Beach and Gulf of Georgia culture types and the historic (post 1853) era.

As indicated in the permit application, the following goals and objectives were accordingly outlined for the project:

1.  To increase the sample of material from the Locarno Beach culture type.

2.  To determine what is contained in deposits underlying those examined last season. In 1994, no excavation unit reached sterile deposits.

3.  To expand excavation in the area containing what is believed to be a house platform dating from the late prehistoric period. The objective would be to uncover details of structure and layout.

4.  To increase the collection of materials relating to the historic period with the objective of adding to our knowledge of how the site has been used since its 1850 acquisition as a school property.

5.  As in the previous year, to provide training in excavation techniques for a University of Victoria archaeological field school. Eighteen students have enrolled and are to be trained in all common aspects of excavation technique and procedure.

6.  For a second season, to provide an opportunity for members of the Archaeological Society of British Columbia's Victoria chapter to participate in a supervised excavation.

7.  The Heritage Properties Branch arranged a scaled-down version of last year's public program. It consisted of tours for schoolchildren, booked in advance.

Archaeological Investigation

To build on the earlier results, 15 of the 1994 units were reopened and 9 new units judgmentally placed to expand excavations in selected areas. By this process, two 2 x 2 m squares examined what in 1994 were seen to be the Locarno Beach culture deposits and a 2 x 3 m excavation explored the possible, more recent house depression.

Control System: The base line dividing east from west lies along the base of the concrete retaining wall at the eastern margin of Admiral's Road between the Bridge and the school house. The north/south dividing line passes through a point that is 11 meters south of a prominent expansion joint in the retaining wall, at the entrance to the bridge. The wall also angles slightly to the west from this point south. A 2 x 2 inch wooden stake is set in the ground at the base of the joint at Nllm/E0m. All grid measurements for the portion of the site investigated by this project lie north and east of the 0/0 reference point. Vertical control is provided by an arbitrary datum plane whose level is equivalent to the highest point on the boulder that forms the body of the historical marker situated by Admiral's Road, a short distance north of the school.

Excavation Standards: All excavation was by trowel by 5 cm level and all material was water-screened through aluminum fly-screening. Units were back-filled after first placing plastic sheeting on the floor.

Recording and Collecting Standards: Three-dimensional point records were kept of all artifacts, whether pre-contact or historic, and of all flaked stone detritus. Provenience by 5 cm level was recorded for all bone, all antler, and selected shell faunal remains. Fire-cracked rock was tallied by count by level. Stratigraphic profiles were produced for all four walls of each free-standing unit. Where units were adjacent, the common wall was not recorded unless, at conclusion of the excavation, one unit was deeper than the other. Sketch floor plans were made in unit field notebooks whenever the base of a level disclosed differences in the distribution of deposits. More formal floor plans were drawn of any features encountered. Most features were also photographed. For reopened, 1994 units, a 20 x 20 cm column of deposit was collected level by level from the corner beginning at the highest elevation. For newly established units, the column sample was increased to 30 x 30 cm, as for these units, the column sample was also to be the sample used for recovery of small fauna.

In 1994, all material retained by the fine screen was bagged for later sorting in the laboratory. While a practical solution for removal of what had proven to be a field-sorting bottleneck, the resulting mountain of level bags was not much reduced by the winter-long efforts of 4 - 5 part-time lab assistants and several volunteers. It was clear that repeating this collecting procedure would only worsen the back-log of unsorted faunal material. To maintain comparability of results, this standard was retained for the 8 units reopened by the field school members. But for those reassigned to the Archaeological Society (6 units) and for any new ones (9 units), collection standards were changed. The column sample was increased to 30 x 30 cm (it remained at 20 x 20 cm for the others) and screened deposit passed by the 5-strand to the inch mesh but stopped by the fly-screening, was simply examined for artifacts.

As in the preceding year, each day's collection of data (including that gathered by the ASBC the previous evening or weekend) was taken to the Archaeology Laboratory at the University of Victoria. On the succeeding day, artifact bags, level bags, screening bags, and column samples were checked for errors in labeling or identification, and that afternoon or the next working day at the latest, discrepancies were resolved with the excavators. This procedure significantly reduced the incidence of residual recording errors.

Artifacts and Lithic Detritus: The number of excavated artifacts catalogued for the 1995 season was 843, of which 647 were historic and 196 prehistoric. In addition, 2 items were recovered from the surface: a flaked slate knife and a ground slate knife. Stone detritus totalled 381 pieces from excavation and an additional 6 from surface collecting--mainly on the beach. Tables 1 and 2 record the numbers of each class of artifact and stone detritus and indicate the distribution of those that came from excavation. For this report excavation units are coded by capitalized letters or letters and numbers. These are also used in Table 3 (which identifies grid location). The information in Table 1 and Table 2 is based on a provisional classification, using mainly field identifications.

Table 1

Table 2

One difference between the 1994 and 1995 reports lies in the way the fragments of ground slate have been treated. In 1994 they were all listed as prehistoric (although the possibility all or some were historic was raised). After this season's work, we feel it more reasonable to consider most of them historic. They come from the same levels as the other historic items and one has marks on it that are presumably from a slate pencil. Accordingly, all ground slate fragments are this year tallied as writing slate and listed with the historic assemblage.

Since the 1994 report was compiled, an additional 52 historic and three prehistoric items have been recovered from analysis of the level bags. The site totals to date are therefore 1546 historic artifacts, 353 prehistoric artifacts, and 470 pieces of stone detritus. These totals take into account the shift of 10 fragments of ground slate from the prehistoric assemblage to the historic assemblage, as discussed above.

Table 3. DcRu 4: Excavation Units and Artifact Densities
(Depth in meters; volume in cubic meters; density in artifacts/cubic meter)

Table 3

Table 3 presents historic and prehistoric artifact densities for each excavation unit along with summary figures in the relevant columns. The volume of sediment within which historic materials were found was calculated for each excavation unit using the minimum depth required to include all discovered historic items. As prehistoric items were found throughout the deposits, the volume of sediment in this case was the total excavated for each unit. The density of historic artifacts is high -- 223 items per cubic meter of historic bearing deposit--while there is a much lower density of prehistoric items -- 14 per cubic meter. If stone detritus is included in the calculation, the density of prehistoric items rises to 38/m3. The density of prehistoric artifacts is identical to that calculated last year, the historic figure is somewhat higher (176 in 1994). In calculating the density of historic artifacts, two in Unit A and two in Unit G3 were eliminated from the tally. In both cases these are units which had been dug to well below the disturbed historic deposits the preceding year. We have assumed the few historic items recovered from these excavation units have fallen from higher up in the walls.

Features: Table 4 lists all features identified during the 1995 season. Most were hearths--simple spreads of ash, charcoal, burned shell, and scorched sediment that define the locations of fires. They are of varying thickness and area and in almost all cases extended beyond the confines of the excavation units in which they were reported. The thickest deposits were usually layered, consisting of a succession of hearths representing repeated use of the location.

A more complex hearth feature is Number 23. It is situated in a 50 cm deep pit which was in part sectioned by the excavation of unit N41-42/E32-33. Sloping layers of ash, charcoal and fire-cracked rock fill the pit. Feature 22 may be a part of this apparent roasting pit as it lies across the top, its saucer shape conforming to the mouth of the pit.

Feature 13, the wooden post in N23-24/E23-24, aligns with similar features discovered the previous season in units coded on the site map as I and K2, forming what appears to be an historic fence line.

Faunal Remains: The arrangement for analysis of the faunal remains is the same this year as last. Dr. Kathlyn Stewart will be accomplishing this at the Canadian Museum of Nature in Ottawa. Highlights of this year's collection of faunal material included continuing

Table 4. Recorded Features for DcRu 4, 1995 Excavations
Table 4

prominence of anchovy and herring, the addition of three-spined stickleback remains, recovery of the first sea mammal specimen (sea lion), and discovery in two locations (Kl and F3 in Fig. 1) of modest quantities of antler "shavings." These curled fragments may have been detached by adzing or carving.

Chronology: One sample collected during the 1994 excavations, and three from 1995 were submitted for analysis in the late summer of 1995. All were of charcoal or charred wood. The results are presented in Table 5.

Table 5. Radiocarbon Age Estimates for DcRu4

Lab Number Weight (mg) Age (years BP) Calibrated Age 95.5% Confidence
R1994-05 TO-5363 423 1070±70 AD 990 AD 855 - 1050
R1995-06 TO-5364 333 2010±50 5 AD 115 BC - AD 90
R1995-12 TO-5365 243 2310±60 390 BC 415 - 200 BC
R1995-18 TO-5366 339 1760±50 295 AD AD 135 - 415

There are three equiprobable calibrated solutions to the age of sample TO-5366: AD 255, AD 295, and AD 320. The middle value is listed in Table 5. The calibration results are those reported by the IsoTrace Radiocarbon Laboratory, using R. P. Beukens' calibration program, C14CAL.

Figure 1

Some comments on the individual estimates follow:

TO-5363: The sample, collected in 1994, comes from unit G3 (Fig. 1) from a hearth feature that was believed to lie on a now-buried house floor. Overlying material contains both prehistoric and historic artifacts, but from the hearth on down, only prehistoric items were encountered. The calibrated age estimate (ca. AD 990) is in accord with expectations, and falls well within the range normally found for Gulf of Georgia culture type assemblages.

TO-5366: From the same excavation unit as the above sample, but recovered in 1995 from the base of cultural deposits a few centimeters above the underlying clay, this was thought to date the Locarno Beach component. At calibrated ages of AD 255, 295, or 310, it is late for that culture type--near the end of what is commonly held to be the period when Marpole culture was dominant in the Strait of Georgia area.

TO-5364: Collected from unit P2, in an area and at a depth considered to be "pure" Locarno Beach culture type, the estimate (calibrated at ca. AD 5) is surprisingly young. It falls at about the middle of what is usually regarded the span of the Marpole culture.

TO-5365: There is little to distinguish the small assemblage of prehistoric items from unit W, but the presence of a microblade very similar in form to those found elsewhere at the site, encouraged a "best guess" of the Locarno Beach culture type. The calibrated age estimate for a sample recovered from 15 cm above the underlying clay deposit, is ca. 390 BC--on the very margin between the Locarno Beach and Marpole periods.

Interpretation of Results: We entered the field this season with a belief, based on the previous year's results, that at least Locarno Beach, Gulf of Georgia, and historic occupations were represented in the site assemblage. It was also thought that the two prehistoric components were segregated. The situation was somewhat muddied by this season's results and complicated by the radiocarbon assays. Two of the Locarno Beach indicators--microblade technology and faceted ground slate--were found to be more widely distributed and unilaterally barbed points--considered typical of Gulf of Georgia--were also found within the "Locarno" deposits.

Although more detailed analysis of distribution, depth, and association will be necessary, some initial observations do reduce some of the ambiguities, leaving much of our earlier interpretation more or less intact.

1.  All microblades from "pure" prehistoric deposits lie between N44 and N79 (01 and 02 to W on Figure 1). Those from south of this area (in units F1 and H)were in deposits containing mixed historic and prehistoric materials. In the case of Fl, at least, these deposits have been interpreted as late fill of house depressions.

2.  A similar depth and distribution relationship exists for the faceted ground slate. The few from undisturbed prehistoric deposits came from north of N47; those south of this line were in mixed historic/prehistoric contexts.

3.  Pieces equillees have not been considered distinctive of Locarno Beach, but they are commonly found in assemblages earlier that Gulf of Georgia. Their distribution at DcRu 4 is generally confined to the Locarno Beach area and lies between N32 and N46.

4.  All small, composite toggling harpoon valves came from units lying between N23 and N30 (units F3 to J). Such artifacts are commonly associated with Gulf of Georgia assemblages.

5.  Of the unilaterally barbed points, the only one from "pure" Locarno Beach deposits was antler rather than bone. If this one is excepted, the others (all of bone) have a generally "southern" distribution and lie between N24 and N45 (G3 and P2). If the site is to be considered segregated, they do extend into the Locarno Beach area. This will have to be explained.

All of these distributions tend to reinforce the notion that within the areas examined, there is a Gulf of Georgia component in the southern part of the site and a Locarno Beach component to the north. The four radiocarbon samples submitted to the University of Toronto's IsoTrace Laboratory were selected to test this postulated segregation of the site's prehistoric components. The assays (see Chronology, above) gave mixed results. In general, the southern portion of the area examined in 1994 and 1995 was younger than the northern. And the buried housefloor that was the objective of the six F and Gunits, dated, as anticipated, from Gulf of Georgia times. But samples designed to date the Locarno Beach occupation were too young. Two fell well within the period associated with Marpole culture and one, the northernmost, lay at the end of Locarno Beach/beginning of Marpole. For the present, there seems no solution to the puzzle of these anomalous radiocarbon age estimates. Expansion of the "Locarno Beach" assemblage is clearly a priority for future work at the site.

Three units (A, G3, and W) were excavated into the sterile, underlying deposit--clay in all cases. These revealed the depths of cultural deposits to be 1.6 m, 1.55 m, and 1.35 m, respectively. S2 (a re-opened 1994 unit) reached a depth of 1.05 m (still in cultural deposit). None of these excavation units disclosed underlying, "different" components, but it should be noted that in the main part of the "Locarno Beach" area, no unit has yet been taken down to sterile deposit.

The historic assemblage was considerably increased during this season. Nails and glass continue to be the principal items found and it is hoped that, in time, their distribution will yield information about site use. Distinctive categories of artifacts (many of which were identified last season) clearly relate to the old Craigflower school. Additional examples of slate pencils, wood pencil parts, and marbles were collected and the category "jacks" added. As explained above, one decision made this year has been to treat all pieces of flat ground slate recovered as fragments of writing slates.

Remains of several recent wooden posts, the first of which were identified last year in K2, were encountered in the complexes of 3 and 6 units lying to the southeast. They form a rough alignment, paralleling the shoreline, and may indicate presence of a fence.

Field School

The pattern of instruction instituted last season was continued in 1995. Students registered for two courses: Anth 390 (Archaeological Field Techniques) and Anth 449 (Archaeology of the Pacific Northwest). The day began in the old Craigflower schoolhouse with a 50-minute lecture, which comprised the Anth 449 part of the field school. This was followed by 5 hours of excavation in site DcRu 4. Each Wednesday, half of the crew would spend their 5 hours at the UVic Archaeology Laboratory sorting bags of faunal material and attempting some preliminary identifications. The eighteen who registered for the field school remained throughout the six weeks.

Each student was responsible for a l x l meter unit. They assisted in setting datum stakes, laid out or re-established the location of units, excavated, screened their own material, drew floor plans, recorded features, collected radiocarbon samples collected column samples, drew stratigraphic profiles, maintained the unit notebook and at the end, produced a brief written summary of the unit's excavation results and then assisted in refilling the excavations.

Public Programming

The project's public interpretation program was scaled down from last year but it provided adequate opportunity for school and other groups to book guided tours. No record was kept this time of the drop-in visitors, but there were many "regulars" as well as numerous casuals.

Of the 17 participating Archaeological Society of British Columbia members, 7 had been involved the previous year. ASBC members had exclusive responsibility for five excavation units: B, H, I, J, and N. The Society worked in the evenings and on weekends. Three members have continued with the project by assisting with the preliminary sorting of faunal material.


John Adams and the Heritage Properties Branch of the Province of British Columbia's Ministry of Small Business, Tourism and Culture, were again "hosts" of the field school and excavation. The Branch also ran a program of school tours and seconded Julie Williams for this task. And Richard Martin of Heritage Properties staff thoughtfully set up warning barriers to keep after-hours visitors from stumbling into the excavations.

The field school members consisted of Caroline Butler, Melissa Carroll, Susan Colleran, Joanne Cumberland, Peter Dady, Jill Doctoroff, Bill Gubbins, Andrew Hickock, Bruce Hoskins, Bernadette Lousier, Mikki Pelletier, Elliott Reynolds, Stephanie Rohdin, Will Small, Byron Smith, Heidi Stack, Dana Uhl, and Volena Wagner. Quentin Mackie and Becky Wigen shared the tasks of leading and instructing this large group. Participants from the Archaeological Society of British Columbia, Victoria Chapter, a group which was again ably led by John Somogyi, included Christine Bailey, Gerald Baran, Joe Baur, Louise Baur, Terry Boorman, Tom Bown, Jonathon Carruthers, Brenda Clark, Pam Killin, Bill MacLennan, Ken Robertson, Elke Sundstrom, Don Turcotte, J. Warris, Sarah Yuile, and James Zwiers. Three volunteers worked a short while near the end of the field period: Lauren Lupton, Capella Sherwood, and Natalie Shumka.

Terry Clark served as laboratory technician and in this capacity visited the site frequently to clear up recording errors and provide logistical support. As a member of last year's field school, his familiarity with both the project and its potential problems was of great help to the instructors. He also worked on into August, organizing material for storage and analysis and assisting in the preparation of this report. A program of laboratory analysis, based on volunteer assistance, carried on into the summer under the general direction of Becky Wigen. Those participating included Megan Bryden, Zoe Jackson, Bernadette Lousier, Lauren Lupton, Bill MacLennan, Elliott Reynolds, Stephanie Rohdin, Capella Sherwood, Natalie Shumka, Elke Sundstrom, Volena Wagner, and James Zwiers.

Once again we are pleased to acknowledge the support and interest of Chief Andy Thomas of the Esquimalt Nation.


Back Contents Next


Logo Industrial Art Internet Group Ltd.
1999 Design and HTML
Maintained by fishAbility