38 Ontario Archaeology No. 72, 2001

Mourning, Curing, Feasting or Industry? The Interpretation of the Quinte and Perch Lake Mounds*

David A. Robertson

The Middle Woodland burnt stone mounds of Prince Edward County, Ontario, and Jefferson County, New York, form an unusual class of monuments that have defied satisfactory interpretation. They have been identified variously as “burial mounds”, “hut rings”, “sweat lodges”, and the remains of “fire rituals”. Some of this confusion may be attributed to the fact that few of these sites have been examined in great detail. More important, however, is the manner in which these interpretations have been formulated, as they have tended to rely on superficial similarities with other feature types and poorly developed analogy. Through consideration of sites that more closely resemble the burnt stone mounds of the eastern Lake Ontario basin, it is suggested that these mounds are more likely to be the remains of seasonally occupied large-scale cook- ing or food processing sites, although it is possible that in some cases they also acquired symbolic significance within the subsistence-settlement systems of local communities.

In what must be among the earliest published Ultimately, however, Wallbridge’s persistence bore accounts of archaeological excavations undertak- fruit with the discovery of burials and other en in the province of Ontario—if not Canada— “objects of curiosity and art” in one mound. Thomas Wallbridge (1860) provided a summary A few years earlier, and across the eastern end of of the results of trenching through at least five Lake Ontario, Franklin Hough provided a pass- examples of a distinct class of late Middle ing reference to perhaps “half a dozen” similar Woodland mounds found along the shores of the “barrows or tumuli” near the town of Pamelia in Bay of Quinte. Wallbridge was both intrigued by the Perch Lake region of Jefferson County, New these unusual features—of which he estimated York (Hough 1851:106-107). By the turn of the there were about 100 in Prince Edward nineteenth century, further explorations had County—and frustrated by the apparent sparse- increased the number of mounds in the Perch ness of artifacts or burials to be found within Lake area to approximately 200. D.S. Marvin— them. Surely these mounds were related to the on the basis of his own observations, the often “barrows or tumuli described by American conflicting accounts of several other turn-of-the- Antiquarians” (Wallbridge 1860:410), which upon century investigators and the accounts of other exploration had routinely yielded spectacular finds, antiquarians in the eastern states—developed the pointing to the former existence in North America argument that these sites were “hut rings” or of a “civilized” race that was subsequently dispos- “camp bottoms” marking the remains of aban- sessed by lesser peoples (Wallbridge 1860:414, doned lodges (Marvin 1887:59-60). This view 417). Nevertheless, most of the Quinte mounds was soon taken up by William Beauchamp that were opened were found to be annular piles of (1887:113; 1905:7; Beauchamp 1894:542; fire-cracked rock, charcoal and ashy soil that were Marvin 1887:59-60), who, at first, was probably otherwise devoid of any clues as to their makers’ less directly familiar with the features, but identities or intentions (Wallbridge 1860:412-413). achieved greater influence as one of Cyrus

*Editors’ Note: This article is part of a set of papers submitted to a special issue of Ontario Archaeology devoted to presentations at the “The Woodlands Revisited”, the Ontario Archaeological Society’s 29th Annual Symposium held in Peterborough in October 2002. Susan Jamieson and Cath Oberholtzer, the editors of the proceedings volume, have graciously permitted this paper to be published here, in order that the journal continues to appear on a semi-regular basis. The remaining papers will appear in the next issue. Robertson The Interpretation of the Quinte and Perch Lake Mounds 39

Thomas’ Mound Survey agents in New York related. In the one instance among contemporary (Julianne Van Nest, personal communication writers where a specific functional interpretation 2002). is offered (Wright 1999:680), it has been shaped A lengthy period was to pass before another more by debates about another type of feature archaeologist took a concerted interest in these found in the Northeast—the semi-subterranean curious features. Throughout the late 1920s and structure—or by ethnohistoric and ethnographic 1930s, William Ritchie visited a number of the accounts of sweat lodges and their attendant rit- Perch Lake mounds, excavating trenches through ual, than by consideration of the Quinte or Perch several of them (Ritchie 1944:313-318). Ritchie’s Lake mounds themselves. first inclination was to accept the idea that these This sort of interpretation is, of course, exceed- features represented domestic sites (Ritchie ingly common in archaeology. Indeed, it may be 1944:317), yet the lack of typical household waste argued that all interpretation, whether couched in at the sites he had investigated was as puzzling to functionalist, processualist, or post-modernist him as it had been to Beauchamp. While he sug- terms, proceeds on the basis of analogical reason- gested that they may have served as sweat lodges ing (e.g., Wylie 1985; von Gernet 1993). The key, (Ritchie 1944:316), he was on the whole left with however, to developing interpretations that are the impression that “some unfathomable ceremo- plausible and compelling in light of the available nialism lay behind the construction of…the Perch evidence is to ensure that the analogues one choos- Lake structures” (Ritchie 1944:318), a conclusion es are appropriate, in that they fit the observed that he did not alter substantially in subsequent phenomenon that is the subject of investigation. years (e.g., Ritchie 1949, 1950, 1969). Both the similarities and the differences between In the thirty years that have elapsed since subject and analogue in term of form and context Ritchie’s last writings concerning the burnt stone must be taken into consideration. mounds (the Quinte mound burials are only If we look beyond the Northeast, we will find mentioned in passing in his Archaeology of New that there may be more appropriate analogues for York State [Ritchie 1980:235], while the Perch the interpretation of the Quinte and Perch Lake Lake mounds are not discussed at all—a reflec- mounds than have been considered to date. These tion perhaps of his frustration with their are to be found as far afield as Texas, and Atlantic intractable character), there has been little effort Europe. The interpretation of burnt stone mounds to investigate these monuments1 nor, apparently, in these different regions has been the subject of has any serious attention been devoted to their considerable discussion as well. Researchers in interpretation. When discussed in broad cultur- these areas have sought even farther afield for ana- al-historical syntheses of the archaeology of the logues and have contested interpretations that have region, the descriptions of these features repro- been predominant for decades. By a curious coin- duce the words of Wallbridge—or of David cidence, these debates have taken place more or less Boyle, who was his earliest commentator simultaneously, but each has been rehearsed with (1897:90)—verbatim,2 or tend to skirt their dis- little awareness that others are grappling with very tinctive attributes by lumping them with other similar questions. In pursuing the question of the monuments that are better documented, and character and functions of the Quinte and Perch seem to fit more comfortably within the archae- Lake burnt stone mounds, it must be emphasized ological record of the Middle Woodland in the that the intention here is not to disparage those eastern Lake Ontario basin (e.g., Funk who have given their consideration to these fea- 1983:340; Spence et al. 1990:166). While tures, rather it is to examine the contexts in which admittedly handicapped by a lack of much their thoughts were formulated, which in turn detailed information on these sites, each attempt influenced judgements concerning the appropri- to account for them has relied on poorly devel- ateness of interpretative analogues, and to high- oped analogy with other types of monuments or light the need for renewed investigation of these features thought to be culturally or temporally sites.3 40 Ontario Archaeology No. 72, 2001

The Quinte and Perch Lake Mounds: soils are generally shallow and unconsolidated, Form, Structure and Interpretation although along the north shore of the peninsula, in the Ameliasburg area, deeper loam and clay The Investigations loam tills occur. Precambrian exposures are lim- The descriptions of the Quinte and Perch Lake ited to a small hill of granite near Ameliasburg mounds offered by Wallbridge (1860), Marvin (Chapman and Putnam 1984:188). Today the (1887) and Beauchamp (1905) remain the most limestone plains of the region, with their alkaline apt summaries of the essential character of these soils, support stands of maple (Acer sp.), beech features, since, with the exception of Ritchie’s (Fagus grandifolia), hemlock (Tsuga canadensis), sporadic investigations, little attention has been elm (Ulmus americana), and basswood (Tilia devoted to them for the past century. A search of americana) in areas of deeper soils, while rock sites in Prince Edward County registered in the barrens, or upland areas of shallow soils, support archaeological database maintained by the Ontario mixed forests of dry-adapted taxa such as pine Ministry of Culture, for instance, revealed only (Pinus strobus) and red oak (Quercus rubra), one site record for a feature that is undoubtedly white ash (Fraxinus americana), maple, and east- a burnt stone mound: apparently one of those ern white cedar (Thuja occidentalis) (Burgar excavated by Wallbridge in 1859. This site clear- 1993). Wetlands and swamps occupy the many ly was registered on the basis of his report alone. shallow depressions on the rock plains, while It is unclear whether any of these features survive marshes border the shoreline and lagoons of the in the Quinte area, given that less than 40 years bay and the river channels that drain into it after Wallbridge’s work “few, if any traces of the (Chapman and Putnam 1984:189). These wetter mounds” were said to be evident (Boyle 1897:90). sites tend to be dominated by hemlock, yellow Survey in Prince Edward County in the late birch (Betula lutea), black ash (Fraxinus nigra), 1960s failed to document any burnt stone white elm (Ulmus americana), and eastern white mounds (Hakas 1967); either they all had been cedar, with pioneering species of white spruce destroyed by this time, or they simply were not (Picea glauca) and balsam fir (Abies balsamea) recognized. Given their supposed ubiquity, the (Burger 1993). Although there have been fluctu- latter seems more likely. In light of this general ations in forest composition due to climatic neglect, any authors who have discussed these change and regional processes of forest succes- features in recent years have been forced to rely sion, this northern mixed hardwood forest likely heavily on the early texts for details concerning prevailed in the region from circa 4,000 B.C. the form of the mounds and as starting points for until the land clearances of the nineteenth centu- their interpretation. ry (cf. Lovis and MacDonald 1999:145). Wallbridge indicated that the Quinte mounds in In New York, Marvin (1887:58) estimated Prince Edward County extended from Rednerville, that there were over 200 mounds, and in Ameliasburg Township, easterly along the shore Beauchamp (1905) reported upon 54 of these of the narrows between Prince Edward and features and the sites of six others, which had Hastings counties, to Massasauga Point at the east been destroyed. These were found primarily end of the Bay of Quinte (Figure 1). He estimated along the bluffs to the north and east of, and that approximately 100 mounds were located overlooking, Perch Lake and on Linnell’s Island, within this area, as well as on some of the nearby a broad, low promontory within extensive islands. Other examples were thought to be found swamps and bogs lying to the south and south- along the Trent River as well (Wallbridge west of the lake (Figure 2). Perch Lake sits at the 1860:411). foot of a series of terraces cut into the limestone, Prince Edward County constitutes a peninsula dolomite and sandstone bedrock that is mantled formed by a Lindsay Formation limestone plain by shallow, poorly drained, heavy clays (Ritchie that extends into the eastern part of Lake 1969:1). The local physiography, vegetation pat- Ontario. Throughout the majority of the region, tern, and drainage regime, at a gross scale, is, Robertson The Interpretation of the Quinte and Perch Lake Mounds 41

CANADA

Ge o MONMONTTREALA rgia n Bay Lake Huron N KINGSTON

TORONTO Lake Ontario

HAMILTON

BUFFALO

Lake Erie

UNITED STAT

RiceRice Lake/Trent ValleValleyy GroupGr

QUINTE GROUP PERCH LAKE GROUP

LakeLakeL a k e OntarioOOntarion t a r i o

0 100 km Figure 1. The location of the Quinte and Perch Lake mounds, together with the better known burial mounds of Rice Lake and the Trent River valley. therefore, broadly similar to that found in Prince the early descriptions note that there was a ten- Edward County. dency for the mounds to occur in pairs: some so The distributions of these features within the closely spaced that they were conjoined; others Quinte and Perch Lake areas are not particularly separated by distances of up to 30 metres. In well documented in the early accounts, nor have such pairings, the two features were said to be of later treatments served to clarify this situation. similar size. Some appear to have occupied hilltops or high The descriptions of the construction of the bluffs; others were apparently located on slopes, mounds provided by Wallbridge and Beauchamp or on the shores of creeks, marshes or lakes in agree that the typical form (Figures 3 and 4) con- low-lying, poorly drained settings. In the Quinte sisted of a ring or “truncated cone” formed of area, and to a lesser extent around Perch Lake, burnt and fire-cracked granite rocks and cobbles 42 Ontario Archaeology No. 72, 2001

Figure 2. Beauchamp’s maps of the distribution of burnt stone mounds in the immediate vicinity of Perch Lake. Adapted from Beauchamp (1905:Plates 1 and 2).

of varying sizes within a matrix of charcoal and ence of rectangular “stone fireplaces” or stone-lined ash-rich soil. These annular features measured “boxes” located in the approximate centres of the roughly 30 to 90 feet (9.2-27.5 metres) in diam- mounds. These features (Figure 5) were dug into eter at the base and rose to a height of five to the original ground surface and lined with flat lime- eight feet (1.5-2.4 metres). The diameter at the stone slabs (Beauchamp 1905:13-14; Wallbridge summit measured approximately 12 feet (3.7 1860:412). Wallbridge and Beauchamp also noted metres), while the central depressions measured that the burnt rock comprising the surrounding approximately eight feet (2.4 metres) in diameter ring of the mounds generally consisted of igneous and were of various depths relative to the sur- materials that were less abundant or easily found rounding banks (Beauchamp 1905:7; Marvin than the locally predominant limestones.4 1887:58-59; Wallbridge 1860:411). Marvin’s Beauchamp (1905:10) took this to mean that the description, however, does not seem to suggest changes limestone undergoes when subjected to that burnt stone was a particularly noteworthy burning were seen to be “objectionable” on the component of the mound banks. part of the builders of these monuments. A number of the reports of the early excavations The general absence of artifacts from either the within these mounds frequently mention the pres- base of the mounds, their burnt rock banks, or— Robertson The Interpretation of the Quinte and Perch Lake Mounds 43

Figure 3. Two mounds on Hyde Creek north of Perch a Lake as sketched by William Beauchamp (1905:Plates 4 and 6). A schematic internal plan of the Figure 3a mound is provided in Figure 5.

b

Figure 4. Cross-section of a Quinte mound excavated by Wallbridge at Massasauga Point apparently—from their immediate surroundings “sepulchral chamber” of a mound at Massasauga was a further source of consternation. On what Point (Figures 4 and 5). These bodies were must have been at least his fifth excavation, accompanied by grave goods that included a uni- Wallbridge finally discovered the remains of five to laterally barbed point, a bone awl, tubular beads of twelve bodies in the central stone-lined box or conch columella, and an incised antler comb 44 Ontario Archaeology No. 72, 2001

Wallbridge's 1860 plan of the stone slab-lined "sepulchural chamber" in the burnt stone mound excavated at Massasauga Point.

Beauchamp's 1905 plan of a burnt stone mound on Hyde Creek showing an idealized rectangular slab-lined "fireplace" in the centre of the feature (D). A, B, and C denote the outer slope, crown of the ridge and the inner slope of the mound respectively. This mound is the same as that illustrated in Figure 3a.

Beauchamp's 1905 plan and cross-section of a burnt stone mound on Cattaraugus Creek. F denotes a stone fireplace. Figure 5. Plans of mound interiors containing central box-like features. From Wallbridge (1860: Figure 2) and Beauchamp (1905: Plate 11 Figure 2 and Plate 12 Figure 2). (Wallbridge 1860:415-417). The comb is sugges- fied that he had proved that the Quinte monu- tive of a late Middle Woodland Point Peninsula ments “belong to the class of sepulchral mounds, complex provenience, circa A.D. 500-700 (Spence such as the observations of Drake, Squier, et al. 1990:166). Although Wallbridge could not Schoolcraft, and many other writers, show to date these finds with any precision, he was satis- exist over a very wide range of country” Robertson The Interpretation of the Quinte and Perch Lake Mounds 45

(Wallbridge 1860:414). While the Quinte mounds as no surprise that Ritchie provided greater detail had largely failed to yield the spectacular finds (Figure 6), although he too found it difficult to reported elsewhere in the eastern part of the conti- explain them. On the basis of his personal expe- nent, they appeared to be the product of the dis- rience in excavating these mounds, he discount- possessed race of Moundbuilders (Wallbridge ed the existence of a subsurface stone-lined pit or 1860:413-414). fireplace (Ritchie 1969:5). Instead, he docu- Initially Beauchamp (1894:540-541) accepted mented the presence of a series of successive that the Quinte mounds were burial features, but in hearths within the central depression of several later years (1905:11) was inclined to believe that mounds, at the base of the peripheral banks and Wallbridge’s burials were intrusive and of no “high at various levels within the fired rock, ash and antiquity”. He likewise came to view accounts of charcoal matrix of the rings (Ritchie 1944:315; human remains found in the Perch Lake mounds 1969:8). He was also inclined to regard Wall- with some skepticism. Marvin’s (1887) suggestion bridge’s multiple burials as “intrusive” within an that the Perch Lake mounds were the remains of earlier monument (Ritchie 1969:8), even though “hut rings” or “camp bottoms”, which was based on he himself encountered “primary” burials in two the work of antiquarians, ethnographers and of the mounds he investigated during his own anthropologists in various areas of the New World, work at Massasauga Point5. In the one instance, a was far more acceptable. These features were most child and an adult female were buried under rock likely the “foundations of early [Algonquian] slab coverings at the base of the central depres- lodges” (Beauchamp 1905:5, 7, 19-20, 22). Ethno- sion of a mound, while in the other an elderly, historic accounts of Algonquian prescriptions possibly male, individual had been interred with- regarding the disposal of waste, particularly with in the rubble deposits of the annular ring respect to the great care accorded to the remains of (Ritchie 1950:259, 261). None of the interments game, were thought to provide the explanation for was accompanied by formal grave goods; howev- the comparative dearth of artifacts or food debris er, random inclusions of corded rocker-stamp on these sites (Beauchamp 1905:22-23). decorated Vinette 2 pottery were found near the Following in Beauchamp’s footsteps in the child burial (Ritchie 1950:259, 1969:9; Spence 1920s and 1930s, William Ritchie visited those et al. 1990:166). A mid-seventh century A.D. Perch Lake mounds that could be relocated and radiocarbon date from one of the Perch Lake carried out limited excavations at a few of them mounds, selected from a series of both earlier and (Ritchie 1944:313-318). He continued to exca- later returns, was cited as evidence confirming a vate such sites, both in New York and at late Middle Woodland, “Kipp Island phase” Massasauga Point, over the next several decades attribution for the basic class of monument. (Ritchie 1949; 1969). In describing the salient Ritchie’s first inclination was to accept the idea features of the mounds in terms of stratigraphy that these features were domestic sites (Ritchie and morphology that he encountered, it comes 1944:317), although such an interpretation was

1 1 1 3

2 1

2 2 2 2 2 1 2 1 3 2 1 1 1 1 1 1 1 1 1 1 1 Forest Duff 1 1 2 2 2 2 1 2 2 2 2 Dark brown sandy soil 2 2 2 2 1 2 2 2 2 Black soil 2 2 2 1 2 2 2 Dark brown ashy soil 1 Limestone (burnt) 2 1 Fire-reddened soil 2 Fired crystalline rock Orange ash 3 Sandstone (burnt) 0 2.0 Charcoal Tan clay-loam subsoil metres Figure 6. Cross-section of a burnt stone mound at Perch Lake through the centre of which Ritchie excavated a two to four foot (0.6- 1.2m) wide trench in 1968. After Ritchie (1969: Figure 1). 46 Ontario Archaeology No. 72, 2001 hindered by the persistent lack of household ings built to maintain the structural integrity of debris, despite the fact that the prevailing soil the mound (e.g., Barber 1976:20). Walter chemistry resulted in optimal conditions for Kenyon, in Mounds of Sacred Earth, which serves preservation of comparatively fragile material as a summary account of his work on mound such as bone. The only substantial quantities of sites throughout Ontario, was not inclined to artifacts recovered from any of the sites he dug offer any definitive statements concerning the were clearly deposited some time after the “anomalous” Quinte and Perch Lake mounds: mounds themselves had been formed. These “Our only achievement so far has been to trans- consisted of scatters of Late Woodland Iroquoian late our ignorance into Latin: we now refer to debris on the upper surfaces of a few sites them as ‘annular’ mounds” (Kenyon 1986:9). In (Ritchie 1944:316). He was compelled, there- their synthesis of the Middle Woodland period fore, to reject Beauchamp’s suggestion that the in southern Ontario, Spence et al. (1990:166) mounds were the remains of earth lodges note only that the Quinte mounds are construct- (Ritchie 1969:8). While he suggested at one ed “largely of stone and [are] frequently without point that they may have served as sweat lodges burials.” They do so in the context of their dis- (Ritchie 1944:316), he was on the whole left cussion of mortuary patterns, implying that with the impression that “some unfathomable monuments of this time period are necessarily ceremonialism lay behind the construction of… associated with funereal rites. The most recent the Perch Lake structures” (Ritchie 1944:318). discussion of the burnt stone mounds appears in Over the next two decades, Ritchie remained of the second volume of J.V. Wright’s mammoth the belief that the majority of the burnt mounds History of the Native People of Canada (Wright represented the remains of late Middle Woodland 1999), and resurrects Ritchie’s sweat lodge inter- ceremonial activity, even if he vacillated over the pretation. As Ritchie had done in 1944, Wright precise form that this ritual took. His last pub- (1999:680) emphasized the possibility that these lished words on the subject suggested that the features were situated in areas of perceived super- Perch Lake mounds, and perhaps some of the natural power or sacred significance that were Quinte features, were the product of an “annual visited on a regular basis, thereby accounting for or cyclical observance…[that was] perhaps some the accumulation of massive quantities of burnt regional expression of new fire cermonialism” debris. (Ritchie 1969:10). He regarded the few Quinte mounds in which he had found “primary” buri- The Interpretation of the Quinte and Perch als as a separate category of monument, which Lake Mounds in Historical Context were related most closely to the Rice Lake burial mounds (Ritchie 1950:261). All archaeological interpretation is shaped by the cultural and intellectual context within which The Syntheses archaeologists find themselves. If additional Given that no published accounts of the excava- demonstration of the truth of this axiom is tion of the burnt stone mounds have appeared required, then it is provided amply by the chang- since 1969,6 subsequent discussions of these fea- ing perspectives on the Quinte and Perch Lake tures have relied heavily on the writings of mounds. Wallbridge, Beauchamp and Ritchie. These later In the 1860s, Wallbridge saw these features as treatments, when they have not misinterpreted testimony to the former presence of a race of the early accounts in terms of the basic structure of sophisticated Moundbuilders in North America the features, have similarly failed to reach a con- who were dispossessed by the natives encoun- sensus as to their function. The central depressions tered by European missionaries and explorers in have been interpreted, naturally enough, as the the seventeenth century (cf. Silverberg 1968; result of looting, while the masses of rock debris Trigger 1989:104-108, 119-126), although he have been re-characterized as exterior stone fac- was apparently more willing than many to recog- Robertson The Interpretation of the Quinte and Perch Lake Mounds 47 nize that the basic Moundbuilder hypothesis tions were immediately evident. Given the reli- rested, in large measure, on “mere conjecture” gious and social roles of sweat-bathing among that “should be received with caution” most of the peoples of the New World, the prob- (Wallbridge 1860:414). Even if removed from able antiquity of this practice, and the long- the main arena of the Moundbuilder debate, standing awareness of the importance of this Ontario researchers were not immune to its practice on the part of anthropologists and eth- effects (Vaccarelli 1995). nologists (cf. MacDonald 1988:17), Ritchie’s By the time that Marvin and Beauchamp were (1944:316) initial proposal that the mounds rep- writing on the burnt stone mounds of western resented the by-products of sweat-bathing made New York, the Moundbuilders myth had been sense, even though he seemed less than entirely thoroughly discredited in scholarly circles. While convinced (Ritchie 1944:318). Later on, he still lacking a firm chronological understanding expressed even greater reservations concerning of the mounds, they were prepared to accept the the sweat lodge interpretation; the scale of burn- fact that they were comparatively ancient. Based ing that occurred on these sites led him to con- on similarities with features seen elsewhere, clude that they were the formed as result of an Marvin posited that the mounds were the annual or cyclical rite of some type of fire cere- remains of circular lodges that were the monialism (Ritchie 1969:8, 10). antecedents of the longhouse. The shift to the At the same time, Ritchie’s investigations longhouse marked an evolutionary development occurred while the basic temporal framework from “the lowest to a higher form of barbarism” that underlies the culture-history of the Northeast (Marvin 1887:60). Whether Marvin equated this to this day was coming together. Indeed, his work transition to the hegemony of Iroquoians over was fundamental to this achievement, as he defined Algonquians in western New York is not clear a series of Archaic and Woodland period phases from his writings. Beauchamp, on the other and complexes with some degree of chronologi- hand, clearly believed that the sites were to be cal control. By the time of his last writings on the attributed to Algonquian occupations that pre- subject at the end of the 1960s, he had thor- dated Iroquoian settlement of the region, based oughly rejected the idea that Quinte and Perch not only on perceived similarities to ethnohistor- Lake mounds could be associated with ically-documented Algonquian beliefs with Algonquians. Dating of the mounds to the sev- respect to the disposal of waste, but also on the enth century A.D. (his Kipp Island phase) effec- assumption that Iroquoians were comparatively tively rendered such an association impossible. recent migrants to their contact-period home- An Algonquin occupation of the eastern Lake lands. These interpretations were significant, Ontario basin at this time no longer fit the pre- however, in that they advocated that the process- vailing cultural-historical paradigm for the es that led to the formation of these features were region demanded by the new in situ hypothesis those of prosaic domesticity. They were the of Iroquoian origins (Ritchie and MacNeish remains of houses rather than of some arcane 1949; MacNeish 1952). Kipp Island was seen as series of mortuary or ritual acts. Such an inter- the direct precursor to the vaguely defined pretation befitted those who were intimately Hunter’s Home phase, which was itself devised familiar with—and contributing to—the grow- to mark the transition between Point Peninsula ing body of research that was firmly grounded in and Owasco (Ritchie 1980:253-254). Late a functional approach to archaeology at the turn Middle Woodland and early Late Woodland thus of the century. represented an unbroken continuum of cultural Over the course of his 30 or 40 years of spo- development in the Iroquoian homelands of the radic investigation of the burnt stone mounds, Northeast. Ritchie, unlike Beauchamp, never really deviated When Ritchie ceased to write on the subject of from the belief that these features were the result the burnt stone mounds, however, the momen- of some form of ritual activity, as no secular func- tum was lost. These features came to be treated 48 Ontario Archaeology No. 72, 2001 in summary fashion as part of a “generic” pack- J.V. Wright’s recent discussion of the burnt age of Middle Woodland cultural traits, even if stone mounds, re-asserts their distinctive quali- they were inexplicable. With its origins in the ties and combines the increasing sedentism and Late Archaic and Early Woodland periods, the association with specific sacred or socially signif- elaborate mortuary ceremonialism of the Middle icant places apparent in the Middle Woodland Woodland—which included the development of with the documented importance of sweating large cemeteries and the use of prominent natural and its attendant ritual during the Late features and artificial mounds—is generally seen Woodland and contact periods. Much as Ritchie as a reflection of the emergence of an increasing- had done in 1944, Wright (1999:680) empha- ly strong sense of social or community identity. sizes the possibility that these features were situ- The long-term use of formalized cemeteries, in ated in areas of perceived supernatural power or some instances including monumental construc- sacred significance that were visited on a regular tion, along with a general increase in sedentism basis, thereby accounting for the accumulation during the Middle Woodland are now thought to of massive quantities of burnt debris. Wright’s point to some important changes in land use and interpretation, however, appears to owe much to control brought about by increasingly sedentary another more contemporary debate in the archae- subsistence-settlement patterns within smaller, ology of southern Ontario, that surrounding the more well-defined band territories (Ferris and identification and interpretation of features Spence 1995:98; Spence et al. 1984; Spence et al. known as semi-subterranean structures. 1990:165-168). In Ontario, these trends are Since they were first identified in the mid- most apparent in the Rice Lake region, a short 1980s as a distinct feature type on Ontario distance to the west of the Bay of Quinte (Figure Iroquoian sites (MacDonald 1986:41-55), semi- 1). Emphasizing ancestral rights to—and impos- subterranean structures have been recognized ing symbols of permanence upon—the land with increasing frequency throughout southern through mound construction became increasingly Ontario. These structures are rectanguloid fea- important as population levels rose and band ter- tures with a lobed projection that together form ritories became increasingly “packed” within a cir- a keyhole-shaped plan (Figure 7). The lobed cumscribed area of rich natural resources. extension normally comprises a ramped entrance Spence et al.’s (1990:166) brief discussion of leading down to the main body of the feature, the Quinte and Perch Lake mounds is clearly the floor of which typically consists of a veneer of premised by this basic understanding of monu- highly organic soil mixed with ash, charcoal and ment construction and burial ritual as an expres- fired soil. Fire-cracked rock is often found in sion of territorial and band identity. Yet by gloss- abundance at the floor level and in the upper fill ing over the distinctiveness of the burnt stone layers, although the latter are not necessarily mounds vis-à-vis the better documented earthen associated directly with the use of the structure. burial mounds of the Rice Lake and upper Trent Post moulds are found around the edge of the fea- River areas (Figure 1), not to mention the appar- ture at or beneath the floor level. It is thought that ent numerical superiority of the former, they these posts represent the vertical elements of the have conflated two distinct monument types. frame of a superstructure covered by bark, skins, The fact that burials have been encountered in a earth, or a combination of these materials (Dodd few of the burnt stone mounds, regardless of and Riddell 1995:149; MacDonald 1988:Figure 2, whether they constitute primary or secondary 1992:329). interments, would certainly suggest that at times The most convincing explanation offered as to some of these features played roles similar to the function of these structures—in that it is the other mounds or places that were resonant with only carefully considered exploration of either meaning, but unlike those found on the shores of structural analogues or northeastern North Rice Lake and the Trent River, the majority of the American aboriginal cultural practice—is that Quinte mounds do not seem to have been so used. they were communal sweat lodges used for ritual, Robertson The Interpretation of the Quinte and Perch Lake Mounds 49

Figure 7. The excavation of an Iroquoian semi-subter- ranean structure. The upper fill layers have been removed to expose the basal floor layer and surrounding ring of struc- tural posts, which are marked by the straws. The ramped entrance to the structure is to the left. Photo courtesy of Archaeological Services Inc.

curative, or socio-political purposes (MacDonald cultural features, regardless of their function, had 1988, 1992; Smith 1976), although uses for gone unrecognized during the previous 20 years other purposes requiring solitude or segregation of large-scale Iroquoian settlement excavations. It cannot be ruled out. While semi-subterranean was also a debate that Wright apparently fol- sweat lodges are apparently a post-thirteenth cen- lowed with interest and which influenced his tury A.D. phenomenon in Ontario (Ferris and thinking on more than just the Quinte and Perch Spence 1995:113; MacDonald and Williamson Lake mounds. In the first volume of his History, 2001:71-72), their precursors are found in for instance, Wright suggested that a series of Pennsylvania and south-western New York as large stratified pits that he and Robert Dailey had early as the tenth century (Smith 1976; found at the Middle to Late Archaic Malcolm MacDonald 1988:21). The frequency with site near Cornwall and had interpreted as earth which these structures occur within longhouses ovens (Dailey and Wright 1955:22), were more on Ontario Iroquoian settlements between circa likely to represent sweat lodges or sweat baths A.D. 1200 and 1500 suggests that their role may (Wright 1995: 250, 252 [Figure 32]). This re- have been a fundamental aspect of daily life in an interpretation was based on a few superficial sim- Iroquoian household, especially if their use relat- ilarities between these Archaic pits (Figure 8) and ed to a curing society that functioned as a social- the semi-subterranean structures of the Middle ly unifying institution within the emergent trib- and early Late Iroquoian periods (i.e., their gen- al systems of the Middle and early Late Iroquoian eral size, flat-bottomed profile, multiple fill layers periods (MacDonald and Williamson 2001:71- and an abundance of fire-cracked rock). It 72; Robertson and Williamson 1998:147). ignored several important differences in terms of The debate over Ontario Iroquoian semi-sub- their basic attributes, namely the lack of periph- terranean structures as sweat lodges during the eral posts of an associated superstructure, the fact late 1980s and early 1990s, at the time when that the floors of the pits consisted of a “solid Wright likely was engaged in writing or revising layer of charcoal atop which rested a layer of much of the text that would ultimately appear in severely burnt and fragmented bone” (Wright the first two volumes of his History, was a con- 1995:250 [emphasis added]) as opposed to a tentious one, spurred on not only by discussion more heterogeneous midden-like accumulation of their appropriate interpretation (for a review of found on the floors of Iroquoian semi-subter- which see MacDonald and Williamson [2001:66- ranean structures, and the fact that “fire-baked 68]), but also by an unspoken concern that these soil lined the sides and bottoms of the features” 50 Ontario Archaeology No. 72, 2001

Figure 8. Profile of one of the Malcolm site features original- ly interpreted as earth ovens (Daily and Wright 1955), but subsequently reinterpreted as a sweat lodge (Wright 1995). Adapted from Wright (1995: Figure 32).

urbed brown topsoil Solid bed of charcoal

ched silty yellow clay Fire-baked clay zone around charcoal bed

Baked grey clay with leached charcoal Sterile subsoil 01.0 Fire-cracked rock

metre (Wright 1995: 250, 252 [Figure 32]). The sparked similar arguments and present similar Malcolm site features, as described by Wright, uncertainties. By coincidence, these debates were are more likely to have formed as a result of in rehearsed at the same time that the sweat lodge situ burning in a sealed, reducing atmosphere debate was influencing Ontario researchers. They that produced temperatures sufficiently high to provide some new alternatives for consideration alter the chemical composition of the subsoil as well. Whether they provide the answers must into which the pit was cut. Such temperatures await further detailed investigation of the Quinte must have been beyond that which could be and Perch Lake mounds. borne by anyone seeking the benefits of the sweat lodge rite. Thus, Dailey and Wright’s original Analogues to the Quinte and Perch Lake 1955 suggestion that the Malcolm site pits repre- Mounds and Analogous Debates sent earth ovens appears to remain the most fea- sible alternative.7 During the late 1980s and early 1990s, researchers What Wright’s 1995 re-interpretations of the in two different parts of the world, with very dif- Malcolm site features and his discussion of the ferent traditions of archaeological inquiry, took Quinte and Perch Lake mounds do demonstrate is renewed interest in sites that bear a number of the powerful influence that discourse on a partic- similarities to the Quinte and Perch Lake ular problem may have on a much wider range of mounds. In Texas, these sites are known as archaeological issues. Obviously, this has both pos- burned rock middens. In Ireland, they are known itive and negative consequences. In the case of the as fulachata fiadh8 (“outdoor” or “wild cooking Malcolm site features, the influence of the sweat places”). Similar features are also found in the lodge debates of the 1980s and 1990s among English Midlands, the Orkneys, and other parts Ontario researchers appears to have led to an of Atlantic Europe, although they are not as unnecessary and ill-advised re-interpretation. common as in Ireland. In each area, concern Nevertheless, by widening consideration of the arose as to the seeming complacency that had burnt stone mounds of the eastern Lake Ontario developed with respect to these sites, particularly basin to include discourses on comparable features as the evidence supporting the prevalent inter- in other parts of the world, it will become appar- pretations, which are clear from their names, ent that sites that are similar in form have pre- remains ambiguous. These debates have yet to be sented similar challenges to interpretation, resolved, but they serve to highlight alternative Robertson The Interpretation of the Quinte and Perch Lake Mounds 51 ways of thinking about the Ontario and New mounds, or on their uppermost surfaces York sites. (Howard 1991:59-63). The absence of such fea- tures within the middle levels of the mounds Texan Burned Rock Middens would appear to be a consequence of their dis- Since 1918, approximately 200 burned rock mid- ruption during the course of repeated use dens have been excavated in central and west cen- (Howard 1991:62) as rock was continually tral Texas (Howard 1991:45; Prewitt 1991:26). rearranged to prepare new working surfaces and The middens represent the accumulation of large new debris was added to the existing dumps. quantities of fire-cracked and broken rock that As the name “burned rock midden” suggests, form heaped mounds. There are two basic forms: the earliest research on these sites was premised domed, which tend to occur in areas of on the assumption that these features represent Precambrian bedrock in central Texas (Prewitt the remains of domestic activities. Recent work 1991:25) and annular, which for the most part are has refined the picture somewhat, with the sug- located in limestone upland areas characterized by gestion that the mounds are the by-products of shallow stony soils in the west central part of the large-scale, seasonal processing of foodstuffs state (Creel 1991:37, 40). The annular forms finding greatest favour. Baking or roasting acorns (Figure 9) consist of a well-defined central depres- or acorn meal and other plant foods in earth sion filled with ashy soil and a few rocks, sur- ovens or on stone “griddles”, or boiling acorns in rounded by jumbled masses of burnt rock debris. wooden containers or skin-lined pits in order to Artifacts tend to be sparse on these sites, but it is leach out their tannic acids, find greatest favour, thought that, for the most part, they date to circa although meat roasting has also been suggested. 3,000-300 B.C. (Prewitt 1991:26), a period char- The acorn processing thesis is based on the some- acterized by foraging and hunter-gathering subsis- what complementary distributions of burnt rock tence strategies adapted to the savannah environ- middens and modern oak savannahs (e.g., Creel ment of the region at that time (Collins 1991:2). 1991), a suggestion that has been partially con- Controlled, broad horizontal excavation of firmed through recent analyses of charred botan- these features in recent years, as opposed to ical remains. The latter studies, however, have trenching (the technique that has prevailed in the resulted in the identification of a diverse range of investigation of the Quinte and Perch Lake nuts, edible grasses and berries, with few taxa mounds), has revealed order amongst the masses represented consistently from one site to the next of discarded rock. Hearths, in the form of basin- (Howard 1991:63). Regardless of the foods being shaped pits that contain accumulations of ash prepared or refined, it is clear that all of these and fired soils, or more rarely set on prepared processing activities relied on hot-stone cooking pavements defined by rock slabs and cobbles, techniques (whether wet or dry), which suggests have been documented at the base of many that these features have formed as a result of the

HORIZONTAL SCALE Burnt Rock 0 5.0 Overburden & Soil metres "Roasting Slab" Limestone Bedrock VERTICAL EXAGGERATION: 500%

Figure 9. Schematic cross-section of a Texan burned rock midden. Adapted from Shiner and Shiner (1977:Figure 15). 52 Ontario Archaeology No. 72, 2001 construction of successive, intersecting hearths a ruined hut” (O’Kelly 1989:223). These mounds on a stable surface over a long period of time in turn surround and overlie a wood plank- or (Howard 1991:57; Collins 1991:5). stone-lined trough, a clay-lined pit, dug-out tree Many questions remain as to the function of the trunk, or similar large tank cut down to the burned rock middens, but it has been recognized watertable or fed by a channel, in addition to one that far more careful and detailed excavation tech- or more hearths and perhaps other ancillary fea- niques will be necessary if they are to be answered. tures (Figures 11 and 12). Recent work in advance of major infrastructure projects has demonstrated Irish Burnt Mounds the existence of many other fulachta that are little Fulachta fiadh, or burnt mounds, form one of the more than thin spreads of charcoal and burnt soil most common types of field monuments in many lacking any associated mound of spent rock parts of the Irish countryside. Estimates based on (Buckley 1991; McLoughlin 2001). Fulacht exca- surveys up to the late 1980s put the number of vations seldom yield a large quantity of artifacts, sites at over 4,000 (Buckley 1990b:9). The “classic” but extensive programs of radiocarbon dating fulacht fiadh is located near a spring or stream in a indicate that these features span the Early Bronze boggy or marshy area and consists of a circular, oval Age (2,300-1,700 B.C.) to the seventeenth cen- or D-shaped mound of fire-cracked rock and char- tury A.D., although the majority seem to date to coal-rich soils, which may be annular, penannular the second millennium B.C. (Buckley 1991:5). or domed in form (Figure 10). “They are generally Based on a variety of legendary or pseudo-his- so unobtrusive as to go unnoticed or to be mistak- torical accounts found in early Irish literary sources en for a mound of field stones or the foundation of dating from the ninth to eighteenth centuries

N

SITE 3

39

.9

39.1

39.3 39.5 SITE 2 39.7

39.7

39. 39.7 5

39.5 3 9 .3 39.3

3 9 .2

39. 1 . SITE 1 9 1 3 .2 39 39.0 39.4

39.6

8 39. 40.

40 0 .2 39.9 40.3 8 40 39. 4 3 4 .4 9 0.1 . 0 8 .6

39.6 39.9 40.7 39.4 40 2 . . 39.8 39 39. 5 0 7 39. 39. 5

40.3

40 39. .1 6 39.9 39.8 40.2 40.0

39. 40 .4 7

39. 5 39.3 4 39. 0.5 1

Stream

0 10 m

Excavation Area Contour Interval: 10cm

Figure 10. Contour plans of three Irish fulachta fiadh or burnt mounds at Killeens, County Cork, Ireland. Adapted from O’Kelly (1954:Figure 8). Robertson The Interpretation of the Quinte and Perch Lake Mounds 53

Limit of Excavation

PLAN LEGEND Charcoal N Paving Stone Spread

Standing Stone Slab ump Stone D urnt f B it o Timber Plank m Li Roasting A Post Pit

Hearth

Trough

Hearth

Pit B White Str ucture (?) Soil

0 5m

Hearth A Trough Hearth B Broken Burnt Stone

Paving Stones Over Hearth

Burnt Stone & Charcoal Ash & Charcoal

Oak Posts Oak Plank 0 2.5m

Figure 11. The burnt mound site of Ballyvourney I, County Cork, Ireland, the excavation of which led to Michael O’Kelly’s influential cooking experiments. Adapted from O’Kelly (1954:Figures 2 and 3). A.D., it had long been assumed that fulachta hunt (Ó Drisceól 1988:673-674, 1990:157). The fiadh represented the remains of cooking places, general lack of associated domestic debris or sub- where meat was boiled following a successful stantial structural remains on these sites, further 54 Ontario Archaeology No. 72, 2001

Figure 12. Plan and profiles of a burnt mound at Coarhamore, County Kerry, Ireland, showing the central trough and the complexity of debris deposition throughout the surrounding banks. Adapted from Sheehan (1990:Figure 13). seemed to reinforce the suggestion that these sites (Figure 11), in which he demonstrated that the were utilized during hunting expeditions away water in the 450-litre-capacity trough could be from major settlements. This interpretation brought to the boil in 30 to 35 minutes through became firmly entrenched among archaeologists the continuous addition of red-hot rocks and during the 1950s following a series of experi- removal of the cooled stones. A 4.5 kilogram joint ments carried out by Michael O’Kelly after his of mutton was cooked in little more than three and excavations at Ballyvourney I in County Cork a half hours at a rate of 20 minutes per pound Robertson The Interpretation of the Quinte and Perch Lake Mounds 55

(O’Kelly 1954:119-123), which is a timing not far works of Early and Middle Woodland groups in off cooking rates using modern convection ovens. the eastern part of the region. Given the fact that Measurement of the volume of shattered rock pro- the Quinte and Perch Lake burnt stone mounds duced by this process led to the suggestion that the date to the later portion of the Middle Woodland mounds of spent rock found at Ballyvourney I and fall within the general area of burial mound (approximately 27m3) represented the debris of construction, it is perhaps natural that they too about 50 cooking events, which likely took place would be assumed to be the remains of ritual or over a number of years (O’Kelly 1954:122-123). highly symbolic acts. It should be apparent by Subsequent experiments were successful in repro- now, however, that there is a need to reassess this ducing Kelly’s results or improving upon them in assumption, even though the investigation of terms of efficiency, and in demonstrating that such these features has languished for more than thirty activities could indeed yield the patterns observed years, and none of those sites that have been exca- during the excavation of these sites (e.g., Buckley vated have been published in extensive detail. 1990a; Ó Drisceól 1988). It is clear that the Quinte and Perch Lake burnt This general view of fulachta fiadh was essen- stone mounds bear close similarities with sites tially unchallenged until the late 1980s, when found in Texas and Atlantic Europe and debate emerged following a proposal that they undoubtedly elsewhere: namely the massive were more likely to be the remains of steam- or quantities of shattered, burnt rock enclosing a even immersion baths (Barfield and Hodder small area, the presence of hearths, deposits of ash 1987; Barfield 1991). This suggestion was based and charcoal-rich soil, and a general dearth of largely on the dearth of faunal remains on many associated artifacts. Some other parallels with the sites, together with some suggestive descriptions burnt mounds of Ireland and Britain are even in the medieval literary record and ethnographi- more striking, although these must remain only cally documented practices in Ireland and north- subjective impressions until further research is ern Europe. The debate subsequently widened devoted to the Ontario and New York sites. In further, to include suggestions that some fulach- terms of site distribution, there is a possible ten- ta may also have been related to the preliminary dency for the Quinte and Perch Lake sites to be stages of wool production, serving as facilities for located in watery areas (this seems most evident washing, fulling or dying textiles (Jeffrey 1991). in the Perch Lake region where Beauchamp Continued experimental work has demonstrated mapped many mounds in the marshy areas sur- that these functions are also quite feasible (Denvir rounding the lake) just as is the case for the Irish 2000). The outcome of these alternative proposals fulachta fiadh. In terms of site structure, several of is the recognition that fulachta, as a broad class, the Perch Lake mounds and at least one of the must be recognized as subsuming a range of com- Quinte mounds likely possessed more or less cen- plex sites of variable functions, each of which must trally located stone slab-lined “sepulchral cham- be approached with considerable attention to bers” or “fireplaces” that were excavated into the detail during excavation (cf. Connolly 2001). subsoil. Although Ritchie (1969:5) was disin- clined to accept the presence of such features, per- An Alternative Proposal for the Quinte and haps in consequence of his reliance on narrow Perch Lake Mounds trench excavation techniques, they are highly rem- iniscent of the troughs found in the European The degree to which the construction of formal burnt mounds, which clearly served as tanks in monuments that leave durable surface traces in which water was heated. Given such formal simi- the countryside was a common practice in the larities, it is possible that they are the products of precontact history of the southern Great Lakes similar activities. remains unclear. In southern Ontario, evidence of A wide variety of hot stone cooking or food pro- monument building that has survived Euro- cessing techniques involving boiling large quantities Canadian land clearance is largely restricted to the of water have been documented ethnographically 56 Ontario Archaeology No. 72, 2001 throughout the New World (e.g., Keene 1981:74- atively small size of the pots. Vinette 2 vessels, for 75; Sassaman 1993:113-115). Such methods were instance, are unlikely to have exceeded 15 litres utilized in situations of immediate consumption in capacity.9 Organic containers such as baskets or were intended to produce foodtuffs amenable may have been somewhat larger, but were quite to longer term storage. Ceramic vessels, skins, or fragile. If using hot rock methods, greater care baskets have all been reported as suitable for the had to be taken not to allow the rocks to burn preparation of foods such as meat, fish, maize, or through the walls of the container as well.10 On nuts through hot stone cookery (e.g., Fenton [ed.] the other hand, a stone-lined tank dug to the 1908:60). Preparation of nuts such as acorns, in level of the shallow watertable in a marshy set- particular, demanded the use of such techniques, ting, or waterproofed with clay chinking if need as they had to be leached in a boiling ash lye to be be, would permit processing on a much larger made palatable (Waugh 1973:82, 122). Boiling scale. The investment in constructing such a fea- large quantities of water was also essential in the ture would be repaid by its permanence and rendering of oils and fats from nuts, fish, or ani- durability. Minimal effort would be required to mal carcasses. refurbish it upon returning to the site after the Nut-bearing trees would have been compara- absence of a year or two. Even during the post- tively abundant in the drier uplands portions of contact period, when ceramic vessels were much the eastern Lake Ontario basin, and could have larger and more durable, the lack of a portable provided an important and storable source of container does not appear to have been an undue protein and fat. High in calories and rich in oil, hindrance to the preparation of food. Bressani nuts may have been an important dietary supple- noted that Huron travelling parties lacking ket- ment, despite the efforts involved in their pro- tles resorted to the expediency of hot stone cook- cessing (i.e., collection, shelling and preparation ing techniques: “they made a ditch in the earth, [Keene 1981:73-75], the latter task requiring and filled it with water, which they caused to boil access to a suitable source of water, which may by cooling in it a number of stones, first heated have involved additional transportation costs in red-hot for this purpose” (Thwaites 1896- the case of acorns). Fish oil may have been an 1901:38:253). equally important commodity. Until the 1890s, The availability of suitable rock may have been eastern Lake Ontario supported a very large lake a greater constraint, particularly if one was sturgeon (Acipenser fulvescens) population (Scott engaged in more extensive processing than the and Crossman 1973:88). Sturgeon are a fatty fish preparation of a single meal. One of the distinc- which, in precontact times, were likely harvested tive traits of the Quinte and Perch Lake mounds in great numbers during the spring spawning sea- is the apparent predominance of igneous and son. Preserving the catch through drying or metamorphosed cobbles among the burnt debris, smoking would have constituted a major task despite the fact that lime- and sandstones form during the seasonal subsistence round. the geological backbone of both areas. Clearly Extracting oil would also have been an important some energy was expended in collecting this activity, as Sagard noted among the Huron: material from drift or creekbed sources. Such “sometimes they boiled the biggest and fattest… efforts were likely repaid by the durability of such to extract the oil from them, skimming it off the rocks in the face of repeated heating and dousing. top of the boiling mass with a spoon” (Tooker Moreover, cobbles collected from creekbeds or 1964:64). lakeshores would have had the additional advan- While heating water in ceramic containers tage of being washed clean. Experimentation has through use of hot rocks placed within the vessel shown that sedimentary rocks produce far more may have been most efficient in terms of the time waste and are reduced to useless fragments far required to bring a given volume of water to the more quickly than igneous and metamorphic boil, the quantities that could be prepared at any materials (Buckley 1990a:171). It has been sug- one time were necessarily limited by the compar- gested that fulachta in areas of limestone bedrock Robertson The Interpretation of the Quinte and Perch Lake Mounds 57 may be more recognizable as upstanding mounds not rule out the possibility that they also fulfilled than in areas where igneous rocks are more read- symbolic roles. People adapt their environ- ily available, simply because of the durability of ment—both conceptually and physically—in the latter material in the face of repeated heating order to establish their place within their own and dousing (Buckley 1990a:171; Condit 1990:21- understanding of the natural and human order of 22). That a degree of selectivity in favour of the world. In doing so they define “places”, igneous and metamorphosed rock was practised at which assume particular significance in specific the Quinte and Perch Lake sites seems apparent situational contexts (cf. Carmichael 1994:96; from the descriptions of Wallbridge, Beauchamp Hubert 1994:16; Ingold 1993:158-161; Tilley and Ritchie, yet the quantities of spent material 1994:22-23). The fact that human remains were generated at these locales were still massive, sug- recovered from several of the Quinte mounds, gesting that the processing activities taking place suggests that such sites played just as important a were carried out on a large scale or were more or role in defining the community as did formalized less annual events within the overall subsistence cemeteries. A similar overlap between “prosaic” round. subsistence-related activities and “ritualized” The fact that the excavations of the Quinte actions seems to characterize shell middens in and Perch Lake sites failed to yield significant coastal regions. Many of the Late Archaic shell quantities of obvious subsistence remains (as has middens of the southeastern United States tended to be the case with respect to the Texan (Sassaman 1993:62-63)—some of which also burned rock middens and European burnt stone resemble aspects of the Quinte and Perch Lake mounds as well) does not necessarily invalidate mounds in terms of basic layout and form—and the suggestion that they served as food prepara- numerous Woodland period shell middens in tion or processing sites. Hot stone boiling tech- New York and New England (Kerber 1999) niques are less likely to result in the incidental include token deposits of human remains, if not charring of plant remains, which is essential to complete interments, as well as deliberate, struc- their preservation in the archaeological record. tured, animal burials. Indeed, parallels are also to Sturgeon, on the other hand, are a cartilaginous found with the better known mounds of the Rice fish; their bones are poorly preserved under most Lake region as well where subsistence and mor- circumstances. Even the skeletal remains of bony tuary activities overlap. Similarly, the tradition of fish are adversely affected by boiling and do not burial under rock slabs appears to be a long- survive well (Needs-Howarth 1999:16, 18). Finally, standing one, even if of variable occurrence (e.g., rendering animal fats likely entailed pulverizing Cook et al. 1998; Jamieson 1992:78). the larger skeletal elements prior to their final Establishing whether the various burials that processing. Only the most stringent of modern Wallbridge and Richtie found within the sampling and flotation procedures are likely to mounds are “primary” or “secondary” remains an recover the remains of such activities. important task, but it should not obscure the possibility that these sites may represent the ves- Are the Quinte and Perch Lake Mounds tiges of a variety of activities. If it is established “Mere” Subsistence Sites? that these sites are simply prosaic food-process- ing sites and all of the human remains found at The suggestion that the Quinte and Perch Lake them are “secondary”, their significance as mon- burnt stone mounds likely represented cooking umental mounds is in no way diminished. or food-processing sites does not necessarily rule Features such as the burnt stone mounds are out the possibility that they served other roles as obviously human creations, yet they have well. While the view of these burnt stone become incorporated into the surrounding land- mounds as subsistence-related features seems to scape as permanent landmarks. Absorption into most closely echo Marvin and Beauchamp’s “hut nature places them beyond challenge, not only to ring” or “camp bottom” interpretation, it does those living at the time of their construction and 58 Ontario Archaeology No. 72, 2001 use, but to subsequent generations as well. Later human and non-human creatures that dwell generations, regardless of whether they attribute therein and have left traces of their own actions the formation of such a monument to human or and understandings. divine agencies, would have had few alternative Thus, as a broad class of feature, the Quinte and responses to the existence of this artificial, yet Perch Lake burnt stone mounds may subsume the permanent feature within the natural landscape. products of a variety of activities. Answering the On the one hand, they could completely avoid it, question as to whether they were “just” food pro- on the other; they could adapt the site to their cessing sites, hut rings, burial sites, sweat lodges, own needs. While the primary monument may or the remains of some form of fire ceremonialism have lost its original significance, it would have must await further detailed investigation. remained symbolically loaded, even if its struc- Nevertheless, it is clear that they represent a hith- ture or meaning was reshaped to suit contempo- erto oversimplified, if not neglected, aspect of the rary needs (Robertson 1992:30). If, on the other late Middle Woodland subsistence-settlement sys- hand, it is established that in certain cases the tems of the eastern Lake Ontario drainage basin. human burials are “primary”, it may be that the Many aspects of the Middle Woodland period in material used to erect the mound over the buri- this region, and for the balance of the lower Great als was collected from nearby food processing Lakes, remain poorly understood, but in general it sites, where the debris left over from hot stone seems that subsistence-settlement systems during boiling represented a ready supply of building this time were characterized by microbands material. Yet this would not necessarily diminish engaged in annual cycles of movement based on the symbolic significance of this debris, for these the seasonal availability of plants, fish and game heaps of burnt stone would likely have been rec- within a comparatively small area, such as a par- ognized as the remains of an essential activity in ticular drainage basin. Seasonal population amal- the life of the community, whether it took place gamations at locales that were particularly rich in yesterday or generations ago. It would simply resources were important venues at which social represent another form of the re-use, adaptation ties between families could be established or reaf- and reinterpretation seen at other earlier mound firmed. sites such as the Middle Woodland Serpent Such a system of mobility would have required Mounds on Rice Lake, the final form of which a diversity of seasonal processing or industrial were the result of accretion over many years, and sites where specific tasks could be carried out and which attracted later Early Iroquoian ossuary which were undoubtedly equipped with the fix- burial as well (Johnston 1968:8, 20, 48-50). It is tures and caches of materials that were necessary the same process—driven by the same basic to complete the work (e.g., Filios 1999:129-134). logic—that underlies all monument construc- This fact alone may account for the apparent fre- tion, regardless of the scale at which it occurs quency and density of features in the Quinte and (Bradley 1993, 1998; Robertson 1992). Perch Lake areas noted at the turn of the nine- Meanings are attached to and gathered from the teenth century. Likewise, the reports of paired or physical surroundings in which people find conjoined mounds may indicate continuity in the themselves and the structures that they create. To occupation or exploitation of certain locales, as refer to such later use of monuments as simply once a processing site became too choked with “intrusive”—a term which has been applied to debris to operate effectively it was replaced by many of the human remains found in the burnt another nearby. It is also possible that such pairs stone mounds and the earthen mounds of Rice represent the simultaneous activities of different Lake—is to overlook the fact that actions are task groups, or that different, but complementary shaped by the immediate experience of living in, activities were carried out simultaneously by a sin- and moving through, the world and by interac- gle group. The seemingly restricted distribution of tion with its physical features and the other these features within the eastern Ontario basin Robertson The Interpretation of the Quinte and Perch Lake Mounds 59 also begs the question of whether these local pop- 2 Compare, for instance, the descriptions provided by ulations had developed subsistence practices that Kenyon (1986:7-9) and Wright (1999:680), which are were unique to the region, or were guaranteed based exclusively on Boyle’s (1897) précis of greater archaeological visibility due to geological Wallbridge’s article. conditions, later European settlement history of 3 A pilot project of mound survey and test excavation the region, or some other factor. in the Perch Lake area, under the co-direction of In any case, throughout the course of the Julieann Van Nest, of the New York State Museum, annual round, task groups of variable composi- and Elena Ponomarenko, of the Canadian Museum of Civilization, represents the first stage in the process of tion—be they gender- or kinship-based or other- acquiring new and detailed information on these sites. wise defined—would have dispersed to such 4 Again, it should be noted that Marvin’s account dif- locations, investing meaning in these places. fered somewhat. He stated that there was “no distur- Places, the paths that linked them, and the his- bance of the soil below the level of the natural torical and social associations that they recall, [ground] surface” and that such rock as was to be together form the mythologized and anthropo- found consisted of small fragments of local limestone morphicized landscapes (Bender 1992:742) that (Marvin 1887:59). are inextricably tied to ideas of community. They 5 By “primary” burial is meant an interment made served as the maps of a community’s history. The prior to or at the same time as the erection of the Quinte and Perch Lake burnt stone mounds were mound, whereas an “intrusive” or “secondary” burial parts of these maps during the later Middle is one that took place sometime after the mound had Woodland. If we are to read these maps anew, the been constructed. 6 burnt stone mounds must be examined in con- Other expeditions to the Perch Lake mounds were cert with other sites of the period. mounted, but these were desultory and very poorly documented (Lennon 1970:7-8). 7 Acknowledgements. My thanks go to Nick It is interesting to note that Wright did not propose Gromoff, Susan Jamieson, Eva MacDonald, Rob the sweat lodge reinterpretation for a similar feature documented at the Late Archaic Knechtel I site in Pihl, Chris Watts, and especially Julieann Van Bruce County (cf. Wright 1972:5, 1995:243). Nest and Ron Williamson, for their comments 8 Fulachta is the plural form. The singular is fulacht. on earlier versions of this paper, or for their sug- 9 The estimate of a maximum of 15 litres is based on a gestions regarding alternative functions for the complete vessel of classic Middle Woodland form burnt stone mounds. Susan Jamieson and Rob (conoidal body and conical base) but bearing later dec- Pihl kindly provided copies of unpublished orative attributes from the Peace Bridge site in Fort reports on the archaeology of the Quinte area, Erie, Ontario. Carbonized encrustations from the ves- while Julieann Van Nest was exceptionally help- sel interior were A.M.S. dated to 1330±60 B.P. (I- ful with respect to the early New York literature. 5243), calibrating to A.D. 675, with a range of 625- Rob von Bitter provided access to the Ontario 860 at 2-sigma (Pihl 1997:404-405; Robertson et al. Ministry of Culture’s Archaeological Sites 1997:501-502). This vessel, which measures 23.5 cm Database for information on registered archaeo- in diameter at the orifice and 37 cm in height, is like- logical sites in the Quinte region. ly at the upper end of the size range for Middle Woodland (Vinette 2) vessels, if not somewhat larger (Robert Pihl, personal communication 2002). Notes 10 The possibility of hot rocks exploding on contact with water and shattering the container was likely less 1 Richard B. Johnston, of Trent University, may have of a danger. During O’Kelly’s experiments, for example, tested a few of the Quinte mounds (Barber 1976:20), it was noted that while the heated rocks might crack however, this suggestion is not borne out by the avail- upon being immersed, they did not shatter. Rather they able archival evidence and may simply reflect confu- tended to break while being handled, either when being sion with other Middle Woodland mounds (Susan removed from the liquid or when placed back in the Jamieson, personal communication 2002). fire to be reheated (O’Kelly 1953:121). 60 Ontario Archaeology No. 72, 2001

References Cited to Burnt Mound Archaeology, edited by V. Buckley, pp 168-174. Wordwell, Dublin. Barber, R.J. 1990b Foreword. In Burnt Offerings: International 1976 Excavations at the Curtis Mounds Site, Contributions to Burnt Mound Archaeology, Southeastern Ontario, 1976. Ms. on file, edited by V. Buckley, pp. 9-12. Wordwell, Ontario Ministry of Culture, Toronto. Dublin. Barfield, L.H. 1991 Irish Fulachta Fiadh: An Overview. In Burnt 1991 Hot Stones: Hot Food or Hot Baths? In Mounds and Hot Stone Technology, edited by Burnt Mounds and Hot Stone Technology, M.A. Hodder and L.H. Barfield, pp. 3-10. edited by M.A. Hodder and L.H. Barfield, Sandwell Metropolitan Borough Council, pp. 59-68. Sandwell Metropolitan Borough West Bromwich, U.K. Council, West Bromwich, U.K. Burger, D. Barfield, L. , and M. Hodder 1993 Revised Site Regions of Ontario: Concepts, 1987 Burnt Mounds as Saunas, and the Prehistory Methodology and Utility. Forest Research of Bathing. Antiquity 61:370-379. Report 129. Ontario Forest Research Beauchamp, W.M. Institute, Sault Ste. Marie, Ontario. 1887 The Aborigines. Traces of a People Long Carmichael, D.L. Since Departed. Transactions of the Jefferson 1994 Places of Power: Mescalero Apache Sacred County Historical Society 1:105-131. Sites and Sensitive Areas. In Sacred Sites, 1894 The Huron-Iroquois District. In Report on Sacred Places, edited by D.L. Carmichael, J. the Mound Explorations of the Bureau of Hubert, B. Reeves and A. Schanche, pp. 89- Ethnology, by C. Thomas, pp. 540-550. 98. Routledge, New York. Twelfth Annual Report of the Bureau of Chapman, L.J., and D.F. Putnam Ethnology to the Secretary of the Smithsonian 1984 The Physiography of Southern Ontario. 3rd ed. Institution 1890-1891. Government Printing Ontario Geological Survey Special Volume Office, Washington D.C. 2. Ontario Ministry of Natural Resources, 1905 Perch Lake Mounds, with Notes on Other New Toronto. York Mounds, and Some Accounts of Indian Collins, M.B. Trails. New York State Museum Bulletin 87, 1991 Thoughts on Future Investigation of Burned Archaeology 10. New York State Education Rock Middens. In The Burned Rock Middens Department, Albany, New York. of Texas: An Archaeological Symposium, edited Bender, B. by T.R. Hester, pp. 1-24. Studies in 1992 Theorising Landscapes and the Prehistoric Archaeology 13. Texas Archaeological Research Landscapes of Stonehenge. Man (N.S.) Laboratory, University of Texas, Austin. 27:735-755. Condit, T. Boyle, D. 1990 Preliminary Observations on the 1897 Ontario Mounds. Annual Archaeological Distribution of Fulachta Fiadh in County Report 1896-97, Being Part of Appendix to Kilkenny. In Burnt Offerings: International the Report of the Minister of Education, Contributions to Burnt Mound Archaeology, Ontario, pp. 89-91. Warwick Brothers and edited by V. Buckley, pp. 18-23. Wordwell, Rutter, Toronto. Dublin. Bradley, R. Connolly, M. 1993 Altering the Earth. Monograph Series 8. 2001 Hot Rocks and Ritual. Archaeology Ireland Society of Antiquaries of Scotland, 15(4):10-13. Edingurgh. Creel, D.G. 1998 The Significance of Monuments: On the 1991 Assessing the Relationship Between Burned Shaping of Human Experience in Neolithic Rock Midden Distribution and Archaic and Bronze Age Europe. Routledge, London. Subsistence in West Central Texas. In The Buckley, V. Burned Rock Middens of Texas: An 1990a Experiments Using a Reconstructed Fulacht Archaeological Symposium, edited by T.R. with a Variety of Rock Types: Implications Hester, pp. 33-44. Studies in Archaeology 13. for the Petro-morphology of Fulachta Fiadh. Texas Archaeological Research Laboratory, In Burnt Offerings: International Contributions University of Texas, Austin. Robertson The Interpretation of the Quinte and Perch Lake Mounds 61

Cook, M., R.G. Mayer, and J. Wilson Hough, F.B. 1998 Osteological Analysis of an Early Woodland 1851 Notices of Ancient Remains of Art in Burial: Mo Pierce Site (AnHq-20). Arch Jefferson and St. Lawrence Counties. Fourth Notes New Series 3(3):13-18. Annual Report of the Regents of the University Dailey, R.C., and J.V. Wright on the Condition of he State Cabinet of Natural 1955 The Malcolm Site: A Late Stage of the History, and the Historical and Antiquarian Middle Point Peninsula Culture in Eastern Collection Annexed Thereto, pp. 103-109. Ontario. Transactions of the Royal Canadian Charles van Benthuyen, Albany, New York. Institute 31(1):3-23. Hubert, J. Denvir, A.-M. 1994 Sacred Beliefs and Beliefs of Sacredness. In 2000 Fulachta Fiadh–An Irish Mystery. http:// Sacred Sites, Sacred Places, edited by D.L. www.angelfire.com/fl/burntmounds. Carmichael, J. Hubert, B. Reeves and A. Dodd, C.F., and D.R. Riddell Schanche, pp. 9-19. Routledge, New York. 1995 The Day Site: A Late Prehistoric Neutral Ingold, T. Cabin Site, Waterloo County, Ontario. In 1993 The Temporality of Landscape. World MTO Contributions to the Archaeology of the Archaeology 25(2):152-174. Late Woodland Period in Southwestern Ontario: Jamieson, S.M. Small Sites Investigations, edited by P.A. 1992 Regional Interaction and Ontario Iroquois Lennox, pp. 189-250. Research Report 24. Evolution. Canadian Journal of Archaeology London Museum of Archaeology, University 16:70-88. of Western Ontario, London, Ontario. Jeffrey, S. Fenton, W.N. (editor) 1991 Burnt Mounds, Fulling and Early Textiles? In 1908 Joseph François Lafitau: Customs of the Burnt Mounds and Hot Stone Technology, edit- American Indians Compared with Customs of ed by M.A. Hodder and L.H. Barfield, pp. Primitive Times. Originally published 1724. 97-108. Sandwell Metropolitan Borough Champlain Society, Toronto. Council, West Bromwich, U.K. Ferris, N., and M.W. Spence Johnston, R.B. 1995 The Woodland Traditions in Southern 1968 The Archaeology of the Serpent Mounds Site. Ontario. Revista de Arqueologia Americana Art and Archaeology, Occasional Paper 10. 9:83-138. Royal Ontario Museum, Toronto. Filios, E. Keene, A.S. 1999 Critical Theory in the Backwater of New 1981 Prehistoric Foraging in a Temperate Forest: A England: Retelling the Third Millennium. Linear Programming Model. Academic Press, In The Archaeological Northeast, edited by Toronto. M.A. Levine, K.E. Sassaman and M.S. Kenyon, W.A. Nassaney, pp. 125-138. Bergin and Garvey, 1986 Mounds of Sacred Earth: Burial Mounds of Westport, Connecticut. Ontario. Archaeology Monograph 9. Royal Funk, R.E. Ontario Museum, Toronto. 1983 The Northeastern United States. In Ancient Kerber, J.E. North Americans, edited by J.D. Jennings, 1999 Interpreting Diverse Marine Shell Deposits pp. 303-372. W.H. Freeman and Company, of the Woodland Period in New England New York. and New York: Interrelationships Among Hakas, T.K. Subsistence, Symbolism and Ceremonialism. 1967 Trent Waterway Archaeological Survey Report Northeast Anthropology 57:57-68. Period 1967. Ms. on file, Department of Lennon, H.A. Anthropology, Trent University, Peterborough, 1970 Mounds at Perch Lake, Jefferson County by Ontario. D.S. Marvin, Esq., Paper Read Before the Howard, M.A. Historical Society, March 15, 1887. Bulletin 1991 Burned Rock Midden Excavations, Hearths of the Jefferson County Historical Society and Botanical Remains. In The Burned Rock 11(2):3-8. Middens of Texas: An Archaeological Symposium, Lovis, W.A., and R.I. MacDonald edited by T.R. Hester, pp. 45-70. Studies in 1999 Archaeological Implications of Great Lakes Archaeology 13. Texas Archaeological Research Paleoecology at the Regional Scale. In Laboratory, University of Texas, Austin. Taming the Taxonomy: Toward a New 62 Ontario Archaeology No. 72, 2001

Understanding of Great Lakes Archaeology, Pihl, R.H. edited by R.F. Williamson and C.M. Watts, 1997 Ceramic Analysis. In In the Shadow of the pp. 125-150. Eastendbooks, Toronto. Bridge: The Archaeology of the Peace Bridge MacDonald, R.I. Site, 1994-1996 Investigations, edited by R.F. 1986 The Coleman Site (AiHd-7): A Late Iroquoian Williamson and R.I. MacDonald, pp. 387- Village in the Waterloo Region. Unpublished 409. Occasional Publications 1. Archaeological MA thesis, Department of Anthropology, Services Inc., Toronto. Trent University, Peterborough, Ontario. Prewitt, E.R. 1988 Ontario Iroquoian Sweat Lodges. Ontario 1991 Burned Rock Middens: A Summary of Archaeology 47: 17-26. Previous Investigations and Interpretations. 1992 Ontario Iroquoian Semi-subterranean Sweat In The Burned Rock Middens of Texas: An Lodges. In Ancient Images, Ancient Thought: Archaeological Symposium, edited by T.R. The Archaeology of Ideology, edited by A.S. Hester, pp. 25-32. Studies in Archaeology Goldsmith, S. Garvie, D. Selin and J. Smith, 13. Texas Archaeological Research Laboratory, pp. 323-330. Proceedings of the 23rd University of Texas, Austin. Annual Chacmool Conference. University Ritchie, W.A. of Calgary, Calgary. 1944 The Pre-Iroquoian Occupations of New York MacDonald, R.I., and R.F. Williamson State. Rochester Museum of Arts and 2001 Sweat Lodges and Solidarity: The Archaeology Sciences, Memoir 1. Rochester, New York. of the Hubbert Site. Ontario Archaeology 1949 Report of Progress. Yearbook 1948: 205-207. 71:29-78. The American Philosophical Society, McLoughlin, C. Philadelphia. 2001 Rathmore Burnt Mound Complex. 1950 Archaeological Investigations in Southern Archaeology Ireland 15(4):4. Ontario, Canada, to Seek the Earlier Stages MacNeish, R.S. of the Owasco Culture. Yearbook, 1949:258- 1952 Ontario Iroquois Pottery Types. National 261. The American Philosophical Society, Museum of Canada, Bulletin 124. Department Philadelphia. of Resources and Development, Ottawa. 1969 Perch Lake Mounds. The Bulletin, New York Marvin, D.S. State Archaeological Association 46:1-10. 1887 Mounds at Perch Lake, Jefferson County. 1980 The Archaeology of New York State. Revised Transactions of the Jefferson County Historical ed. Harbor Hill Books, Harrison, New York. Society 1:58-61. Ritchie, W.A., and R.S. MacNeish Needs-Howarth, S. 1949 The Pre-Iroquoian Pottery of New York 1999 Native Fishing in the Great Lakes: A State. American Antiquity 15(2):97-134. Multidisciplinary Approach to Zooarchaeological Robertson, D.A. Remains from Precontact Iroquoian Villages near 1992 The Navan Forty Metre Structure: Some Lake Simcoe, Ontario. Unpublished PhD dis- Observations Regarding the Social Context sertation, Biologisch-Archaeologisch Instituut, of an Iron Age Monument. Emania: Bulletin Rijksuniversiteiet, Gronigen, Netherlands. of the Navan Research Group 10:25-32. Ó Drisceól, D. Robertson, D.A., and R.F. Williamson 1988 Burnt Mounds: Cooking or Bathing? Antiquity 1998 The Archaeology of the Parsons Site: 62:671-680. Summary and Conclusions. Ontario 1990 Fulachta Fiadh: The Value of Early Irish Archaeology 65/66:146-150. Literature. In Burnt Offerings: International Robertson, D.A., R.F. Williamson, R.I. MacDonald, Contributions to Burnt Mound Archaeology, R.H. Pihl, and M.S. Cooper edited by V. Buckley, pp. 157-164. Wordwell, 1997 Interpretations and Conclusions. In In the Dublin. Shadow of the Bridge: The Archaeology of the O’Kelly, M.J. Peace Bridge Site, 1994-1996 Investigations, 1954 Excavations and Experiments in Irish edited by R.F. Williamson and R.I. MacDonald, Cooking-places. Journal of the Royal Society pp. 493-510. Occasional Publications 1. of Antiquaries of Ireland 84:105-155. Archaeological Services Inc., Toronto. 1989 Early Ireland: An Introduction to Irish Prehistory. Sassaman, K.E. Cambridge University Press, New York. 1993 Early Pottery in the Southeast: Tradition and Robertson The Interpretation of the Quinte and Perch Lake Mounds 63

Innovation in Cooking Technology. University 1649. Bureau of American Ethnology of Alabama Press, Tuscaloosa. Bulletin 190. Smithsonian Institution, Scott, W.B., and E.J. Crossman Washington, D.C. 1973 Freshwater Fishes of Canada. Bulletin 184. Trigger, B.G. Fisheries Research Board of Canada, Ottawa. 1989 A History of Archaeological Thought. Sheehan, J. Cambridge University Press, New York. 1990 The Excavation of a Fulacht Fiadh at Vaccarelli, V. Coarhamore, Valentia Island, Co. Kerry. In 1995 Nineteenth Century Archaeological Thought Burnt Offerings: International Contributions in Ontario: David Boyle and the Mound to Burnt Mound Archaeology, edited by V. Builder Myth. In Origins of the People of the Buckley, pp. 27-37. Wordwell, Dublin. Longhouse: Proceedings of the 21st Annual Shiner, J.L., and M.V. Shiner Symposium of the Ontario Archaeological 1977 Structure and Organization of a Burned Society, edited by A. Beckerman and G. Rock Midden: The Indian Creek Site, Real Warrick, pp. 129-139. Ontario Archaeological County, Texas. Plains Anthropologist Society, North York, Ontario. 22(2):263-282. von Gernet, A. Silverberg, R. 1993 The Construction of Prehistoric Ideation. 1968 The Mound Builders. Ohio University Press, Cambridge Journal of Archaeology 3(1):67-81. Athens, Ohio. Wallbridge, T.C. Smith, I.F. III 1860 On Some Ancient Mounds upon the Shores 1976 A Functional Interpretation of Keyhole of the Bay of Quinte. The Canadian Journal Structures in the Northeast. Pennsylvania 29:408-417. Archaeologist 46(1-2): 1-12. Waugh, F.W. Spence, M.W., R.H. Pihl, and J.E. Molto 1973 Iroquois Foods and Food Preparation. Facsimile 1984 Hunter-Gatherer Social Group Identification: ed. National Museums of Canada, Ottawa. A Case Study from Middle Woodland Originally published 1916 Geological Survey Southern Ontario. In Exploring the Limits: Memoir 86, Anthropological Series 12. Frontiers and Boundaries in Prehistory, edited Department of Mines, Ottawa. by S.P. De Atley and F.J. Findow, pp. 117- Wright, J.V. 142. International Series 223. British 1972 The Knechtel I Site, Bruce County, Ontario. Archaeological Reports, Oxford. Archaeological Survey of Canada Mercury Spence, M.W., R.H. Pihl, and C. Murphy Series Paper 4. National Museum of Man, 1990 Cultural Complexes of the Early and Middle Ottawa. Woodland Periods. In The Archaeology of 1995 A History of the Native People of Canada, Volume Southern Ontario to A.D. 1650, edited by C.J. 1 (10,000 B.C.-1,000 B.C.). Archaeological Ellis and N. Ferris, pp.171-188. Occasional Survey of Canada Mercury Series Paper 152. Publication 5. London Chapter, Ontario Canadian Museum of Civilization, Ottawa. Archaeological Society. London, Ontario. 1999 A History of the Native People of Canada, Volume Thwaites, R.G. 2 (1,000 B.C.-A.D. 500). Archaeological 1896-1901The Jesuit Relations and Allied Documents. 73 Survey of Canada Mercury Series Paper 152. volumes. Burrows Bros., Cleveland, Ohio. Canadian Museum of Civilization, Ottawa. Tilley, C. Wylie, A. 1994 A Phenomenology of Landscape: Places, Paths 1985 The Reaction Against Analogy. In Advances and Monuments. Berg, Providence. in Archaeological Theory Volume 8, edited by Tooker, E. M.B. Schiffer, pp. 63-111. Academic Press, 1964 An Ethnology of the Huron Indians, 1616- New York.

David A. Robertson Archaeological Services Inc. 528 Bathurst Street, Toronto, Ontario, Canada M5S 2P9 [email protected]