Mercury Series


by Christopher Ellis and D. Brian Deller


This study reports on the information recovered from the Parkhill Early Paleo-Indian (fluted point) site in southwestern Ontario. It is the type site of what has been called the Parkhill complex or phase in the central Great Lakes area. Dating to the end of the last ice age, Parkhill is located just inland from the southern margin of Lake Huron. It represents the activities of the earliest human occupants to enter Ontario as the continental glaciers retreated to the north. Found and investigated in the early 1970s, at the time it was the first major site of this age and affiliation discovered in Ontario and only the second to be discovered in Canada after the Debert site in Nova Scotia reported in the 1960s. Parkhill is a large archaeological site, made up of at least nine concentrations of lithic debris distributed over an area of some 6 ha in cultivated fields. These large sites are the rarest of Paleo-Indian sites and only one other such site is now reported for Ontario. As a result of the above factors, the site was named of National Historic Significance by the Historic Sites and Monuments Board of Canada.

The Parkhill site is situated just north of a modern creek at a major right-angle bend in an abandoned shoreline which represents an early lake level above that of modern Lake Huron. The abandoned lake level has long been believed to have been occupied by an earlier pro-glacial lake at around 11,000 to 10,400 radiocarbon years B.P. and later, after a period of low water, below modern levels, by a post-glacial Nipissing phase lake at ca. 5500-4500 B.P. Parkhill represents only one of at least 17 fluted point sites in the area above and ringing that lakeshore, most of which are assignable to the Parkhill Phase. The fact that no fluted point sites are known below that lake level across all of southern Ontario suggests they were occupied contemporary with the earlier lake level (e.g. in the 11th millenium B.P.) and this age estimate is consistent with radiocarbon dates on such sites throughout North America.

Excavations were carried out at Parkhill primarily in three of the site areas (13, C, and D). The other areas are known solely through surface collection off of ploughed fields in the 1970s. In addition to the excavations, attempts were made to recover information on the geological and paleoenvironmental setting of the site. These investigations occurred both at the site itself and by a series of boreholes taken up the creek valley south of the site. The geological and paleoenviromental studies confirm what has long been suspected: that a lake level above modern existed beside the site during the estimated time of its occupation between ca. 11,000 and 10,400 B.P. Actual dated lake bottom deposits representing this high water level were recovered some 1.2 km up the creek valley south of the site. These data indicate the site was located adjacent to a large estuary of that lake at the time it was occupied. Although such a high water level had long been expected based on data from elsewhere in Ontario to the north, these deposits represent the first confirmed and dated evidence of that lake level (usually referred to as Main Lake Algonquin but perhaps attributable to a post-Algonquin Lake Ardtrea) in the southern Lake Huron basin. Analyses of pollen and other paleoenvironmental evidence such as fossil beetles from the deposits related to this lake, along with evidence from work elsewhere in Ontario, indicate that the site was situated in a spruce-dominated landscape through much of the estimated time of the occupation. Significant amounts of non-arboreal pollen suggest the site was most likely in an open woodland, rather than a closed forest, environment.

Although cultural features were encountered at the site, the main source of information on the inhabitants are the recovered stone artifact assemblages -- all that usually survives on sites of this age. Also, the collections include the largest known samples of certain tool forms yet reported for Ontario sites such as finished fluted points, miniature points made on channel flakes, and trianguloid end scrapers. They represent a major building block in documenting and understanding Paleo-Indian tool kits. Therefore, the stone artifacts, including both tools and debris from manufacture and repair, were subjected to very detailed analyses. Besides summary descriptions of the 330 stone tools and preforms recovered, as an aid to future comparative studies, measurements and attribute states for each individual item recovered are presented in appendices, and almost the whole assemblage of formed tools is illustrated.

The artifact descriptions focus on explaining patterned artifact variability in the assemblage in terms of a series of interacting immediate/proximate causes such as raw material selection processes, manufacturing procedures, hafting or non-hafting of implements, resharpening and so on. The main stone material used at the site is Collingwood chert from the Fossil Hill formation. This material, which is the dominant one employed at all Parkhill Phase fluted point sites in the area, had to be procured and brought to the site from locations 175-200 km to the northeast in southcentral Ontario. Smaller amounts of material from other locations, notably Bayport chert from the vicinity of Saginaw Bay in Michigan, attest to interaction with areas farther afield. Detailed models of primary core reduction procedures are developed which suggest the inhabitants used standardized core forms of both a conical and bifacial shape. Blanks selected from the reduction products were made into a very wide variety of implements, most of which are highly retouched. The assemblage not only provides documentation of some well-known artifact forms, such as fluted points, but also provides information on some other much rarer and less widely recognized tool forms, such as large bifaces resharpened by alternate edge bevelling, backed bifaces, hafted perforators, large concave side scrapers with spurs, and large parallel-sided end scrapers. The almost 6000 pieces of flaking debris recovered are notable for their small size, with only 29 pieces weighing over 1 g each. These data indicate that primary reduction activities were minimized. One presumes the earlier wasteful stages of core reduction were carried out only near toolstone sources to maximize the portability and usability of material to be transported some distance. A notable - component of the debris assemblage is a good sample of almost 200 channel flakes derived in point fluting. These flakes, and the recovered preforms, allow the suggestion that failure rates in the difficult fluting operation were rather low and probably less than 10-15% at this site.

Detailed information is provided on the spatial distributions of artifacts, debris and features in the three main excavated areas B, C and D. Certain patterns seem evident in these data such as a tendency for debris to concentrate around features and for tools such as piercers or gravers to occur in multiples, perhaps because they had short application lives such that more than one tool was needed to carry out a particular task.

An examination of the overall arrangement of all Parkhill site areas, suggests that it is unlike most of the previously reported large, multi-locus, Great Lakes/Northeast, Paleo-Indian sites such as Fisher (Ontario), Nobles Pond (Ohio), Bull Brook (Massachusetts) and Vail (Maine). The distribution of different artifact and debris forms by Parkhill area, clearly allows one to recognize two groups of those areas. One group, consisting of areas A/J, B, C and E, is notable for having the highest percentage of fluted bifaces ever reported from the over 50 fluted point sites or site areas now known in the central Great Lakes. They are very specialized work areas of a relatively small size (ca. 200 to 400 mz) where fluted biface rehafting (as indicated by discarded point bases/ears) and manufacture (as indicated by channel flakes) was emphasized almost to the exclusion of all else. All of these site areas are in a linear array along the western site margin, literally paralleling the adjacent lake shore and looking across the former narrow lake estuary to the south/southwest. The second group, consisting of the remaining five site areas (D, G, H, I, K), are irregularly distributed across the eastern half of the site away from the lakeshore, and are quite different. Minimal excavation and small sample sizes places some constraints on interpretations, but these areas have more diverse tool inventories and were the scene of a wider array of domestic activities with unifacial tool use being more common. These areas seem to vary in size but most are much larger than the western site areas, and the most intensively investigated (Area D) exceeds 2000 m2. Since Area D alone is almost three times the size of the largest previously reported whole site in southwestern Ontario (Thedford II), a relatively large occupying group is suggested.

Eastern Area D can be linked to one of the western site areas (Area B) by artifact crossmends suggesting they were used at the same time. A parsimonious interpretation of the site as whole, reinforced by areal variation in raw material frequencies and in point morphology, is that it was used on several occasions and that each occupation may have involved use of one eastern general domestic activity locus and one western specialized fluted point maintenance and production area. The western areas are situated lining the strandline with a good view along the shoreline to the west and north. These may have been ideal areas to watch for game approaching along the lakeshore while preparing or maintaining weaponry (hence, their quite specialized assemblages) while the eastern areas were actual domestic camp locations. Caribou have long been suggested to have been a favourite target of Great Lakes Paleo-Indians based on environmental reconstructions and recoveries of highly fragmentary remains of that species at other sites such as Udora (Ontario) and Holcombe (Michigan). Parkhill was located just to the north of a narrow estuary in a situation identical to that employed to hunt caribou by modern hunters. These locations are ideal for intercepting caribou as the animals are slowed and easier to harvest while crossing such waters. As such, one can suggest that the Parkhill site as a whole seems to fit the stereotype often applied to all large PaleoIndian sites: an aggregation site where people seasonally gathered for the intercept communal hunting of game such as caribou.

Return to my place in the Mercuries List

ASC HistoryASC ArchaeologistsCurrent Research Projects Upcoming Field ProjectsRecent Field ReportsASC Publications ASC Electronic PublicationsCanadian Archaeological Radiocarbon Dates Virtual ASC Archaeological ExhibitsRevista de Arqueologia Americana