THE MARTIN SITE OF GRAND ISLAND: THE ARCHAIC AND WOODLAND OCCUPATION OF WESTERN NEW YORK Ezra B. W. Zubrow and Paul T. Buerger Department of Anthropology University at Buffalo INTRODUCTION This paper describes the Martin site and its relationship to the prehistory of the Niagara Frontier. It focuses on site description, chronology, and function. Over 140,000 artifacts were recovered from this location although only two percent of the total area was excavated. There was the equivalent of 1500 large waterpots and five and a half tons of lithic material. A large amount of faunal refuse, especially fish completes the site. Location The Martin Site is located on the southeastern shore of Grand Island, New York. It is twelve miles upstream from the great cataract of the Niagara River. The Problem This paper addresses two questions. First, when was the site occupied and second, what type of site occupation was it? The authors show the site was occupied intermittently from 3500 B.C. to A.D. 1600 when it suddenly was abandoned. Site function changes from "hunting" to "hunting, fishing and habitation" to "hunting, fishing, habitation and agricultural production" and finally reverts to " a hunting site with minimal habitation". These transitions were later than in the Northeast in general. Although the site changed over time there was little technological innovation and thus this paper contends that the Martin site exemplifies "evolution without innovation." History of the Excavations Field work at the Martin site occurred over four field seasons spanning two decades. The early field work was under the direction of Marian White while the later field seasons were conducted by Ezra Zubrow. In 1962, amateur archaeologists excavated an 8'x 7' test pit at the site and concluded from ceramics that it contained evidence of a possible Owasco pre-Iroquoian culture. They reported their work to Marian White and in 1963, she decided to invest a full field season into excavating the site. Marian White divided the site area and the work into three parts corresponding to the general topography. First, there is a marshy strip extending from the river's edge 300 feet. A small excavation called the X - area, revealed a rich midden. Second, above this marshy area there is a terrace about 6 feet high and half a mile long. Just behind this terrace was the original test pit that Marian White expanded into a large scale excavation. From the terrace, the land rises gently for 600 to 700 feet at slightly less than a two percent slope. The western 300 feet were cultivated and are now the 16th and 17th holes of the Beaver Island golf course. Third, Marian White directed her student Joyce Holloway to do an intensive surface survey of the area that is now under the golf course. She found several areas of occupation (Holloway 1964). The published record is scanty. Marian White only published a preliminary article concerning the Martin Site prior to her death (White 1964). Thus, the authors felt it incumbent to review all her old notes and students' papers in order to correlate her data with Ezra Zubrow's. It is important that the information be published and not left to lie unused in the files of archivists. Marian White concluded: (1) There were several distinct occupations located on an east to west line. (2) Several of these occupations were to be destroyed by the golf course. (3) There seems to be a continuous occupation north to south along the edge of the terrace. (4) There are three stratigraphic levels. The upper top soil level divides into a disturbed and non-disturbed strata. The disturbed strata is the upper 0.6 feet where plowing destroyed the context. Below 0.6 feet is the non-disturbed strata and at the base, a clay sub-soil. (5) Type points were found of every major New York State culture from the Lamoka Archaic (3000 B.C.) to the present. Ceramics similar to Owasco pottery included plain and decorated wares. Marian White tentatively dated the primary occupation to the Late-Middle Woodland or Early-Late Woodland, approximately A.D. 1 to A.D. 1,000. (6) A large amount of bone was preserved in the site. Fishing was an important part of the subsistence base as indicated by the large proportion of fish bone and occasional net sinkers. (7) Numerous prehistoric pits contain large amounts of artifacts while post holes indicate structures. In 1978, Zubrow reopened excavations at the Martin Site and continued them in 1979. The new work had several explicit goals. The first was to obtain data relevant to the ecological and biogeographic aspects of island cultural evolution. The second goal was to place the Martin site into its cultural context in relationship to other sites in the immediate area. Third, the aim was to test a series of methodological procedures including nested random surveying and Kriging. The field work in 1978 was divided into 5 major archaeological divisions. First, there was a regional site survey of Grand Island in which a nested shovel test design was used in a series of randomly selected quadrants. Twenty two quadrants were selected and a nested transect pattern of shovel tests ranging from intervals of 3 yards to 200 yards were excavated. Second, a systematic shovel test pattern was laid out across the entire terrace area in order to determine explicitly the size of the site. Third, Marian White's original 1963 excavation was expanded. Fourth, a series of 42 Krig pits were excavated. Kriging is a prospecting strategy which may be used as a search strategy at different levels of resolution. The Krig predictions from the 1978 data, the trend analysis of the 1963 excavation, a trend analysis of the 1978 systematic shovel test survey and a unit excavated at the far southern end of the site all showed a second concentration of material in the southern part of the site. Thus, in 1979 this predicted area was excavated using 2.5'x 5' and 5' x 5' squares also 15 more Krig units. This concentration of excavation in a second area provides a sample for comparison with the 1963 data. In short, the archaeological fieldwork has been at five levels of resolution, regional survey for regional trends, systematic shovel test pits to find the size of the site, Krig tests for predicting patterns within the site, and 5'x 5' squares excavated for detailed area-specific information. The amount of excavation on the Martin Site by year and type is shown in Table 1. IMMMMMMMMMMMMMMMMMMMM; : : : TABLE 1. HERE : : : HMMMMMMMMMMMMMMMMMMMM< THE CHRONOLOGICAL PROBLEM - WHEN WAS THE SITE OCCUPIED? When Marian White published her preliminary report on the Martin Site, she predicted: When our study of the Martin Site material is completed, our knowledge of the culture sequence in the Niagara Frontier will be considerably fuller.... Soon we will be able to describe the local variants of the New York State culture sequence which occurred in the Niagara Frontier (White 1964:54). She noted: "The Martin Site was unique for it is a multicomponent site in which every cultural phase from Lamoka Archaic onward is represented. The only exception is the Niagara Frontier Iroquois (White 1964:51,54)". Yet she underestimated its complexity. The site turned out to be complex spatially, chronologically, and culturally. The Martin site contrasts with the typical prehistoric sites in the region. Most are single component sites. Some of the sites represent the same village that has been relocated. Marian White, studied the characteristics of prehistoric villages and their movements. She found: (1) the average length of occupation was twenty years; (2) the villages moved an average distance of two miles because of firewood and small animal habitat depletion (Hunt 1986:313). Such sites yielded related artifacts from a specific time period or cultural phase, a typical single component site. The Martin site is a major departure from this type of limited occupation because it is a multicomponent site spanning multiple millennia. The contrasting metaphors of "vertical temporal depth" to "horizontal spatial spread" is apt. The Martin site is constricted by its island location and isolated in a manner in which most northeastern prehistoric villages are not. A principal reason for this departure lies in the abundant and renewable source of fish that was a mainstay of the Martin subsistence pattern. Another aspect of the chronological problem is the considerable disturbance. It affects seriation and stratigraphy. Even historic ceramics and glass have been found in the lowest level of excavation. In the early part of this century, two large and deep marine slips were dug into the shore line in the most archaeologically sensitive area of the site. The excavated fill was spread out over a large area. In addition, the nineteenth century home of Captain Martin was located in this area and the remainder of the site was disturbed by plowing. The problem is to separate the artifact material into the appropriate cultural and chronological sequence and to distinguish the multiple occupations. Ideally, there should be an analytical cross section of the site as if it had a single continuous stratigraphic profile. Instead, the multiple components partially overlay each other. Frequently, they are spatially separated; occasionally they are mixed. Analysis of Chronological Data Radiocarbon dates and ceramics span the Woodland Phase, while the projectile points from the site span both the Archaic and Woodland Phases. Dated charcoal from hearths cover a period from 670 B.C. to 1530 A.D., with a grouping of three dates between 255 A.D. and 775 A.D. (see Table II). This time span includes the entire Early, Middle and Late Woodland periods, clustering around the Middle Woodland. IMMMMMMMMMMMMMMMMMMMM; : : : TABLE II. HERE : : : HMMMMMMMMMMMMMMMMMMMM< The first step was to divide the total site artifact assemblage into broad categories, such as; lithics, ceramics, floral and faunal elements (see Table III Artifact Tabulation). The second step was to identify datable objects and arrange them chronologically. Two categories of datable objects were apparent. One was the series of projectile points, which could be arranged by typology into a series of five cultural phases; two Archaic and three Woodland. Table IV shows the entire collection of projectile points from the excavations at the Martin site. Marian White classified the projectile points according to the standard typology as defined by Ritchie into such point types as Lamoka, Brewerton, Meadowood, and Madison (Ritchie 1961). Then, these were grouped chronologically by cultural stage such as Late Archaic, Middle Woodland etc. Years later David Sadow classified the additional points excavated during 1978 and 1979 excavations (Sadow 1980). These have been added to White's projectile point tabulation (see Table IV, Martin Projectile Point Type Classification). IMMMMMMMMMMMMMMMMMMMM; : : : TABLE III. HERE : : : HMMMMMMMMMMMMMMMMMMMM< IMMMMMMMMMMMMMMMMMMMM; : : : TABLE IV. HERE : : : HMMMMMMMMMMMMMMMMMMMM< The other category of datable objects was a series of decorated rimsherds, which represented each of three Woodland cultural phases (see Table V Rimsherd Type classification). By comparing these two sequences across time, points and rimsherds, an initial index of the time of occupation might be derived. IMMMMMMMMMMMMMMMMMMMM; : : : TABLE V. HERE : : : HMMMMMMMMMMMMMMMMMMMM< The projectile point typology provides the longest sequence extending from Late-Middle Archaic (3500 B.C.) to Late Woodland or Owasco pre-Iroquoian (1600 A.D.). The numerical division is one third Archaic points to two thirds Woodland points. More specifically, there was a numerical peak in the Late-Middle Archaic, and later a major increase in the Middle and Late Woodland Phases. (see Table IV). The analysis of rimsherd typology of 167 specimens from the 1963, 1978 and 1979 excavations showed 28 plain sherds, 31 Vinette Type I sherds classified as early Woodland, 99 Point Peninsula type sherds classified as Middle Woodland, and 9 Owasco classified as pre-Iroquoian or Late Woodland cultural phase (Buerger 1992). Numerically for the ceramics there was a single peak in the Middle Woodland. This analysis differs markedly from White who, as mentioned previously, believed that the major pottery type distribution belonged to the Owasco pre-Iroquoian. This ceramic study agrees with Noonan's Master's thesis on the Martin site. She believed the predominant pottery type was Point Peninsula, Hunters Home Phase (Noonan 1966). In none of the ceramic studies has there been evidence of Historic Iroquoian Pottery at the Martin Site. The discrepancy in the stages when ceramics and points rose and fell caused the first interpretative dilemma. One plausible explanation is that the site was used differently at different periods. The point series suggested periods when the function of the site was a hunting camp. The pottery series, on the other hand, represents the periods when the site was used primarily as a habitation site. Another interpretation could be that points and ceramics show differential intensity of occupation. The next analytic effort was to determine the contents of each cultural phase. To do this one has to have clear stratigraphic layers or clear criteria of association. Since stratigraphy was not possible, one turns to the criteria of association. The following analysis makes use of two of these criteria separately and exclusively. The first was to associate the artifact assemblage to datable points (see Table VI). The second was to associate the artifact assemblage to datable ceramics (see Table VII). After each association was made, the distribution was tabulated. IMMMMMMMMMMMMMMMMMMMM; : : : TABLE VI. HERE : : : HMMMMMMMMMMMMMMMMMMMM< IMMMMMMMMMMMMMMMMMMMM; : : : TABLE VII. HERE : : : HMMMMMMMMMMMMMMMMMMMM< The Projectile Point Association, (Table VI), lists three columns per point phase. The first column is the artifact count of each type of artifact associated to that phase as determined by the projectile point chronology. The second column is the first column expressed as a percentage of the total number of that type of artifact in the site without respect to phase. The third column is the average weight of the artifacts in the artifact count category. It is simply the sum of the weights of the artifacts in the artifact count category for that phase divided by the number of the artifacts in the artifact count category for that phase [17]. The second association shown is the association of artifacts to datable ceramics by cultural phase in table VII. Essentially the same procedure that was used for the projectile point association was repeated for each pottery type. All the identifiable ceramics were divided into three types: Vinette Type 1, Point Peninsula, and Owasco pre-Iroquoian. Thus, chronologically, the Vinette Type I pottery is dated to the Early Woodland; the Point Peninsula type pottery is the Middle Woodland; and the Owasco pre-Iroquoian type pottery is the Late Woodland. The spatial units are the units of association. Thus, if the association unit included Vinette Type I pottery, it was dated Early Woodland, similarly if it included Point Peninsula type pottery, it was dated Middle Woodland, and if included the Owasco pre-Iroquoian type pottery, it was dated Late Woodland [18]. Table VII shows three columns for each period per ceramic phase. The first column is the artifact count of each type of artifact associated to that phase as determined by the ceramic chronology. The second column is the artifact count of each type of artifact associated to that phase as determined by the ceramic chronology expressed as a percentage of the total number of that type of artifact in the site without respect to phase. The third column is the average weight of the artifacts in the artifact count category. How do the artifact frequencies of the two associations compare? Several interesting results are apparent. First, there is a general pattern. Artifacts found in association with points stay constant or increase through the Woodland Stage until they reach their peak concentration in the Late Woodland Stage. Artifacts associated with Woodland Stage pottery on the other hand show the greatest concentration in the Middle Woodland Stage. Second, there are no anvil stones, gorgets, bone fragments, or any bone tools associated with projectile points. On the other hand there are no flint needles associated with pottery. There are surprises. One might expect bone fragments and bone tools that are products of hunting to be associated with the tools of the hunt, the projectile point. Yet, this is not so. Similarly, one might expect flint needles that are domestic household tools to occur in association with pottery. This does not occur either. There are two aberrant patterns of artifacts associated with rim sherds. The first is a "flat" pattern across time. Burins and sedimentary rock are the two types of artifacts that make up this pattern. The second is a "decreasing" pattern across time. The artifacts that follow this pattern are polished stone, clam shells, bone tools, and "kaolin" pipe stems [19]. Turning to the Archaic, the only datable association is with the projectile points. There is a general pattern. All artifact categories increase in the Late-Middle Archaic to a primary peak in the Late Archaic. This pattern applies to all artifact types except decortification flakes or ground slate [20]. It also is the pattern of bone. One fact was clear, there was little technological innovation in stone tool technology throughout the Archaic and Woodland phases. The major technological development was in ceramic production where refinement in decoration, the use of temper, and thinner vessel walls increased through time. As an archaeologist, one aspires to an undisturbed excavation unit in which all the cultural phase indicators exist in the proper sequence. This ideal unit could in some sense typify the site. Fortunately, such a unit does exist at the Martin Site. On the northern periphery of the site, about twenty feet from the water line is unit 70'N 20'E. It was excavated in 1978 (see Table VIII Unit Profile). The stratigraphy reveals artifacts that conform to a Middle to Late Woodland Cultural Stage in proper sequence. When the top soil was removed, five features were revealed; two appear to be hearths and the other three are storage or trash pits. Radiocarbon dating from feature 9 revealed two dates; 320 A.D.+\-65 and 525 A.D.+\-140. Both are clearly Middle Woodland. Feature 9 also contained a Middle Woodland Levanna Point. The other hearth, feature 24 had charcoal that tested 1460 A.D.+\-70 indicating Late Woodland. This unit exemplifies the Martin Site where the primary cultural concentration dates to the Middle and Late Woodland Cultural Stage. IMMMMMMMMMMMMMMMMMMMM; : : : TABLE VIII. HERE : : : HMMMMMMMMMMMMMMMMMMMM< TYPE OF SITE - A FUNCTIONAL PHASE BY PHASE ANALYSIS The Martin site was occupied irregularly from 3500 B.C. to 1600 A.D. In the next part of the paper, the authors show the type of site occupation varied from a "hunting site" to a "hunting, fishing and habitation site" to "a hunting, fishing, habitation and agricultural production site" and finally reverts to a "hunting site with minimal habitation". Two areas of the site showed the highest concentration of artifacts. The first was designated as the potential habitation site and was located between 70'N/15'S and 15'E /60'W. A second area of concentration between 265'S-310'S and 65'W-120'W represents a work area. It lacked signs of habitation. If one maps the density of lithics for each category of lithics, all but four types have the highest concentrations of lithics in the habitation area. This spatial distribution is consistent with each nuclear family producing tools at their individual habitation. The area adjacent to the Martin site was a choice hunting and fishing ground. The southern shore of Grand Island between Strawberry and Motor Island has been and continues to be a major fish spawning area in the Upper Niagara River. A series of ecological studies shows 29 species of native mammals (12 herbivorous, 12 carnivorous, 5 omnivorous), 34 types of birds (29 aquatic and 5 terrestrial), and 41 kinds of fish (whose total density has ranged from 0.5 to 3.5 tons per square mile)[21]. The Lake Forest Vegetative Zone characterizes the vegetation on Grand Island and is typical of both the Canadian and Carolinian biotic provinces. There is a microclimatic zonal reversal immediately contiguous to the site. This area is slightly warmer in the winter and cooler in the summer because of the water temperatures. In addition, the complexity of the general biota is increased because of the increased sunlight reaching low level plants. The forest canopy does not obstruct them for they are adjacent to the water. Beaches favor the islands, wet and marshy lands; oaks and maples favor the sandy ridges; and hemlocks the few valleys. Chenopodium, polygonum, amaranth, grapes, berries, and nuts provide potentially gatherable plants (Mellody and Johnston 1978). The long term stability of the Grand Island's boreal forest is somewhat poor since the nutrient reservoir is partially dependent upon a relatively high density of clay. In the area immediately surrounding the Martin site the soils tend to be sandy and well drained resulting in a leaky nutrient reservoir. This is not particularly conducive to agriculture. In contrast, the faunal analysis shows that the hunting of wild game and birds and fishing was a productive occupation of the aboriginal population at the Martin site (see Zubrow 1980, Table IX, Ellis 1985). IMMMMMMMMMMMMMMMMMMMM; : : : TABLE IX. HERE : : : HMMMMMMMMMMMMMMMMMMMM< The approach used is to take the classic model for each cultural period represented at the site and match this with supporting artifactual data. The cultural foundation of the Woodland is the Middle and Late Archaic periods. The focus will be on the Early, Middle and Late Woodland periods. Late Middle Archaic Phase The sequence begins with an initial artifact assemblage of Otter Creek Points from the Late-Middle Archaic Phase. There are other associated artifacts at Martin supporting a Late Middle Archaic occupation. They include small numbers of flakes, retouched flakes, bifaces, points and scrapers. The flakes are less than one percent of the total number of flakes found at Martin. For the rest of these artifact types, the proportions are similar. There is a single large ocher piece which represents six percent of the total ocher found at Martin. In addition, 86 or 1.9% of the total ceramics are intrusive into this phase and should be ignored (Table VI). The area of the Martin site from which Late Middle Archaic artifacts and debitage are found is very small. At best, it represents a few hundred square feet. The artifact assemblage is not sufficient for sophisticated functional interpretation. One might want to argue that the site was used sporadically as a hunting site. There is nothing to support fishing. The authors have created several indices of intensity of occupation which are informative. One index is a temporally weighted number of artifacts per period. It is calculated by taking the number of artifacts per period and adjusting by the length of the period. For the Late Middle Archaic this value is 0.8. A second index is the total number of artifacts per phase divided by the total number of artifacts in all phases as associated by the points. In other words, this is the percentage of total artifact assemblage by point phase. For the Late Middle Archaic this is one percent. Late Archaic and Transitional Phase The Late Archaic at Martin contains many more types of points than Late Middle Archaic. These include Brewerton Side-notched, Corner-Notched and Eared-Triangle, Normanskill, Lamoka and Genesee Points. Susquehanna Broad, Perkiomen Broad and Dry Brook Fishtail points were found in the Transitional Phase at Martin. Fishing was an important part of the Archaic adaptation at the Martin site. The Lamoka Lake site, the type site for Lamoka points, is usually credited with being the first locality where fishing was practiced extensively in New York State (Ritchie 1980:48). The occurrence of net sinkers defines this fishing adaptation. At Martin, 21% of the total number of net sinkers date to this phase. The authors believe that the Late Archaic-Transitional phase at Martin was a habitation and a hunting site. One indicator are multiply reused hearths. Secondarily, although most bone was unclassifiable fish bone, there were more than 2800 bone fragments primarily from deer. There were no intact long bones. Of these 433 or 17% are found in association with the Late Archaic points. All of the bone had been fractured and the absence of trabeculated bone in the interior shows the marrow had been extracted. The authors suggest such large scale extraction of marrow could not take place in a short time or "on the move" and therefore is an indication of habitation. The large amounts of lithic material in various stages of production supports habitation. There is no quarrying but one finds cores, chunks, intermediately finished tools including unifaces and bifaces, finished tools including points and scrapers and large amounts of debitage including slightly more than 17,000 flakes. This suggests a long term occupation of the site. Other evidence is the seasonality of the fauna. There were immature specimens of deer bone, fish bone and seasonally migratory bird bone. These seasonal indicators show that these species were killed in different seasons. Yet taken as a whole all four seasons are represented at Martin. Multiple seasonal use makes multiple seasonal occupation more probable. Finally, one would expect artifacts to be light at non-habitation sites since they must be transportable. Yet, at the Martin site during the Late Archaic and Transitional most of the artifact types do not have the lowest average weights. There are 7 categories of artifacts that have the lowest average weight in the Late Archaic and Transitional compared to 25 categories of artifacts during the Woodland. When the authors did their association analysis, there were a large number of ceramics which were associated with late Archaic points to this phase. Some of them are clearly intrusive such as the historic pottery, "kaolin" pipes and ceramics and can be ignored. However, there are 4,607 prehistoric potsherds making up 24% of the entire prehistoric potsherd collection as well as 167 rimsherds that are 49% of the rimsherd collection associated with late Archaic and Transitional Phase points. The authors believe that the majority of these are not intrusive but are Late Transitional. Examining the artifact distribution from the Late Archaic and Transitional phase, several anomalies have implications for the function of the site. The authors wanted to determine which artifact types might be most significant in each phase. A somewhat arbitrary level of importance was created. If an artifact type had more than 20% of the total number of artifacts of that artifact category in the entire site during a single phase, the authors thought it might be highly relevant for that phase. For the Late Archaic and Transitional phase these "highly relevant" types included for lithics: points (21%), drills (30%), burins (33%), net sinkers (21%), sedimentary rock (47%), ground stone (37%), and polished stone (33%); for ceramics: potsherds (24%), rimsherds (49%), historic pottery (33%), and kaolin stems (46%); and for ecofacts: clam shells (41%) and charcoal (49%). The relative importance of domestic activities at this time are indicated by the large amounts of sedimentary rock that are usually ascribed to stone boiling for cooking and the polished stone that is used in grinding. The occurrence of clam shells in significant numbers is a strong indication of trade with the eastern seaboard. In the Late Archaic there is evidence of connections from the Martin site to other areas. Table X shows the artifacts by period whose point of origin may be determined by unique types of lithic material. The evidence shows that during the Archaic there are examples of lithic material from Diver's Lake (25 miles), Leroy (40 miles), Pennsylvania (225 miles) and Normanskill (300 miles). In other words there are connections to the south and east. IMMMMMMMMMMMMMMMMMMMM; : : : TABLE X. HERE : : : HMMMMMMMMMMMMMMMMMMMM< The intensity of occupation increases during this period. The temporally weighted number of artifacts per period for the Late Archaic and Transitional is 12.4 or more than ten times the previous period. The value of the second index, percentage of total artifact assemblage by point phase, is 34%. Measured either way, there is a significant increase in the intensity of occupation. Early Woodland Phase The Early Woodland period in Western New York is a specific entity, i.e., the Meadowood phase. The diagnostics are the classical corner notched points of the Meadowood type and the Vinette Type I pottery . Both Marian White and the authors identify the Martin site as having a Meadowood component (White 1964:54). Projectile Points of the Meadowood corner notched variety formed 11% of the total number of identified points as did Fulton Turkey Tail 1% (Table IV). Vinette Type I rimsherds formed 22% of the total identified decorated rimsherds (Table V). At Martin this phase is less important than the Late Archaic and Transitional and the Middle Woodland. The total area, the number of artifacts, and the number of artifacts in most tool categories are all smaller than in the "Late Archaic and Transitional". Using the intensity indices, the percentage of total artifact assemblage by point association, is 11%, one third of the previous period's 34%. The percentage of total artifact assemblage by ceramic association is 29%. However, these are somewhat misleading for the Early Woodland is a short period. The temporally weighted number of artifacts per period for this phase continues to increase and is 18.8. Meadowood sites are located characteristically along rivers and small lakes (Ritchie 1980:81). Large deposits of fish refuse have been found in association. The inference is fishing is important in the subsistence pattern. At Martin the amount of fish bone would support this view. However, the number of net sinkers decreases from 21% (11) to 9% (5) during this phase. In addition, the average weight of the net sinkers decrease by one third which the authors claim signifies a change in technique since there are no indications at this time of different current erates in the Niagara River. The Riverhaven II site, located on the east branch of the Niagara River 4 miles downstream from the Martin site, is a classic Meadowood horizon (Granger 1974). In many ways the sites are analogous during this phase - similar riverain location, similar artifact distribution, and similar fauna. One major difference between Martin and the other Meadowood sites is that eight out of ten of the other Meadowood sites in this area were Red Ocher burial sites. No burials were found at Martin nor in the immediate vicinity. Furthermore, there is little red ocher - 2% (9). Since Red Ocher burials have been cited as evidence of ceremonialism at some Early Woodland Sites, Martin does not conform. During this phase there are no "highly relevant" artifact types if one uses the 20% criteria on the artifact distribution associated by points. (Table VI) The pottery style at Martin is similar to the Anderson and Locust Point sites at Selkirk Shores State Park in Oswego County during the Meadowood and the succeeding Point Peninsula phase (Birnie 1989:54, P.C. Hayes 1990). It also shows a close relationship to the Wray and Vinette sites. The datable Vinette I ceramics are 22% or 31 decorated rimsherds (Table V). By point distribution association, there are 1109 prehistoric potsherds (6%) and 11 (3%) rimsherds (Table VI). However, by the ceramic association there are 3,985 (21%) prehistoric potsherds and 235 (67%) rimsherds (Table VII). If one uses a 33% criteria on the artifact distribution from this phase associated by ceramics, then drills, burins, ground stone, ground slate, polished stone, rim sherds, "kaolin" pipe stems, clam shells, and bone tools are "highly relevant". While a specific type of activity is difficult to ascertain from these figures, it may be an indicator that habitation was relatively more important than hunting and fishing during this phase since net sinkers, points, scrapers and other hunting or fishing tools do not reach this level (Table VII). During the Meadowood phase there are close cultural relationships between New York and other states. The closest are to the Red Ocher and Glacial Kame cultures of the Upper Great Lakes region (Ritchie 1980:201). However at the Martin site, the actual range of "the areas of origin" for lithic artifacts is smaller than the range in the Late Archaic (Table X). There is no Pennsylvania flint. Thus, the southern influence is less important. In summary, there is fish bone, trade flint, pottery, and artifacts indicative of a habitation site. The earliest radiocarbon date comes from a hearth dated to this phase. Considerable evidence of hunting and fishing occurs at Martin but with decreasing intensity. During this period, there is no evidence of agriculture, i.e. corn cobs, seeds or pollen. The numbers and weights of the artifacts both decrease from the previous period. In short, occupation intensity has diminished from the previous phase. Middle Woodland Phase The Middle Woodland phase in Western New York and southern Ontario is frequently classified as the Hunters Home- Point Peninsula. Culturally, this phase has the closest affinity with the upper Great Lakes in Central and Western New York and southern Ontario. Characterized by Vinette Type II ware, this ceramic tradition has considerable stylistic variety. The most frequent decorative technique is the impressed cord wrapped stick. It creates a variety of patterns as seen in the Martin collections. At Martin the projectile points from this phase are Jack's Reef Corner Notched and Levanna. They are 5 (3%) and 19 (14%) of identifiable point collection (Table IV). Vinette Type II rimsherds formed 71% of the total identified decorated rimsherds (Table V). At Martin this phase is very important. Using the intensity indices, the percentage of total artifact assemblage by point association, is 28% second only to the Late Archaic and Transitional and by ceramic association 59%. The temporally weighted number of artifacts per period for this phase continues to increase and is 21.64. Marian White believed the ceramics showed a close cultural relationship to southern Ontario (White 1964:54). The analysis by the authors of the Martin ceramics would indicate a closer similarity to the ceramics from Selkirk Shores and Vinette sites. Thus, they are more closely related to sites of Central New York. Using the criteria of "highly relevant" artifact types from the point association distribution, one finds cores, drills, decortification flakes, ground stone, polished stone, rim sherds, "kaolin" pipe stems, clam shells, and ocher pieces. Using the criteria of "highly relevant" from the pottery distribution, retouched flakes, chunks, unifaces, points, drills, choppers, burins, net sinkers, ground stone, ground slate, polished stone, hematite, anvil stone, gorget, potsherds, rimsherds, historic pottery, beaver incisor, clam shell, and bone tools are all above 33%. This is the largest grouping of "highly relevant" tool categories. (Table VII). The significance of this abundant amount of artifactual material would indicate that the Middle Woodland was indeed the peak phase of prehistoric occupation and functionally has all the characteristics of both a hunting and fishing site as well as a habitation site. Most of the radiocarbon dates date to this period (Table II). In addition, most of the hearths are from this time. By far the vast majority of artifacts by number belong to this phase according to the pottery association i.e. 29% Early Woodland, 59% Middle Woodland, 12% Late Woodland. Martin is situated in the center of the Point Peninsula culture which is located in Southern Ontario as well as Western and Central New York. Fishing and the collection of fresh water mussels were the primary industries with hunting of lesser importance. (Ritchie 1980:209). During this phase at Martin, the net sinkers are 52% of the total collection of net sinkers and clam shells were 45% of the entire collection of clam shells. It has generally been assumed that maize agriculture was not practiced in the Middle Woodland phase in this area of the Northeast. At the Martin site there is no evidence of corn cobs or kernels from this period. One of the authors with the archaeometry graduate research group of the University at Buffalo participated in an analysis of the organic contents of encrustation on ceramic sherds from this area. An encrustation sample was obtained from a Middle Woodland sherd from the Martin site. It was one of sixty one samples on which inductively coupled plasma spectrometry (ICAP) analyses were run. The ICAP analysis of this encrustation, nearby agricultural soil, and contemporary maize crops showed a .9620 "r squared" value for the variation among seventeen trace elements. This shows that maize or at least organics with the same trace element distribution were being harvested and consumed in the immediate area. This pattern was also found in the sites of four of the five geological zones which make up the western part of the state. In only the southwestern part of the state near Lake Erie and the Pennsylvania border was this pattern not confirmed (Fie et. al. 1990). In terms of other external patterns, there is no evidence of trade to the South or to the East. Yet, there is evidence of trade to the West, i.e. to and from Ohio. This is indicated by artifacts from Flint Ridge (Table X). In summary, this phase represents the cultural climax at the Martin site. Late Woodland Phase The dates for the late Woodland phase in New York are 1000 to 1600 A.D., and this phase is generally known as the pre-Iroquoian or Owasco. The later portion of this phase is considered to be contemporary with the introduction of maize agriculture in Western New York. The 1460 A.D. radiocarbon date from a hearth at Martin is contemporaneous with other sites in this area which demonstrate maize agriculture by excavated corn cobs. Other hallmarks of this phase are Madison points and Owasco pre-Iroquoian ceramics (Table V). One might expect with agriculture a greater intensity of occupation. However, at Martin the intensity of habitation occupation of the site seems to be declining in this period. The temporally weighted number of artifacts per period for the Late Woodland is 47.15. This is twice the previous period. The value of the second index, the percentage of total artifact assemblage by point phase, is 27%. This in turn is approximately equal to the previous period. If, however, one calculates the percentage of total artifact assemblage for the ceramic association it equals only 12%. The authors believe that one may interpret the first two indices to suggest that hunting continues to be increasingly important but that the third index suggests that habitation is diminishing in significance. The rationale for this interpretation follows. The most dramatic artifactual finding at the Martin Site in the Late Woodland phase is the large number of Madison points ie., 43 as compared to 40 for the Early and Middle Woodland combined. This is even more significant when one considers that approximately half of the Middle Woodland Levanna points should be counted as Late Woodland because their documented range of usage was from 700-1350 A.D. If one assumes an even distribution through time, this is well into the Late Woodland time frame. These findings combined document the continued or even expanded importance of hunting in the Late Woodland phase. The same emphasis cannot be said for the continued evidence of habitation at the peak level noted in the Middle Woodland phase. All indicators including the number of decorated rim sherds, the artifacts associated with pottery and radiocarbon dates have declined in number. The distribution for decorated rimsherds are 22%, 71%, 6% for Early, Middle, and Late Woodland and the association of all artifacts to datable ceramics are analogously 29%, 59% 12%. This would seem to indicate that while the inhabitants continued to hunt the area, their level of habitation is reduced to a level below the Early Woodland and similar to the Late Archaic and Transitional. Using the criteria for "highly relevant" for ceramics, the following tool types have more than 33% of their total amount in this phase. They are burins, ground stones, ground slate, and polished stone. Using the analogous criteria of 20% for the point association, the "highly relevant" types are drills, sedimentary rock, hematite, burnt flint, flint needles, and clam shell. Whether one considers the bone fragments associated by projectile point type or associated by ceramics, the amount of bone fragments is considerably less than the previous period, i.e. a decrease from 14% to 9% or from 25% to 7% respectively. One needs to remember that in this study the authors use bone fragments which have their marrow removed as an indicator of habitation and not primarily hunting. There is little evidence of external trade during this period on the basis of flint. The existing evidence is from the Onondaga deposits of Fort Erie, approximately 10 miles away to the west. This is the first connection seen to Southern Ontario. Yet, the importance of this is really an issue regarding modern boundaries rather than prehistoric ones. On the other hand, the fact that by projectile point association there is a large amount of associated clam shells (64, 50%) meaning trade did exist to the eastern seashore. SUMMARY The Martin site was the largest and longest continuously occupied prehistoric site on the Niagara Frontier. The principal occupation of the Martin site began in the Late Middle Archaic and reached a major peak in the Late Archaic and Transitional and subsided during the Early Woodland period. Reaching a maximum florescence during the Middle and early Late Woodland periods, it went into a sudden decline. Substantiating evidence is based upon artifacts associated with datable points and ceramics as well as five radiocarbon dates that clustered around 500 A.D. Beginning as a part-time fishing and hunting camp in the Late-Middle Archaic phase, it was characterized by Otter Creek Points. The Late Archaic was characterized by a variety projectile points from several different locations to the East and South. During the Early Woodland, identification with other sites of the Meadowood phase on Grand Island and also on the mainland was common. The material manifestation of this cultural expansion was the standardized assemblage of Meadowood corner notched points and the Vinette Type I pottery. The most intense occupation of the Martin site occurred in the Middle Woodland cultural phase during which Point Peninsula type of pottery was produced. There are indications for the beginnings of agriculture based on the organic contents of ceramic encrustation. Expansion continued with the cultural relationships shown throughout Central and Western New York State and also Southern Ontario. Ceramics at Martin conform to the well developed traditions of Central New York. This is anomalous given the historic occupation of the Niagara region by the Neutral Iroquois whose culture centered in Southern Ontario. CONCLUSIONS The Martin site is a multicomponent site on the southern end of Grand Island. The inhabitants were varied with economic and cultural relations shifting over time. It was a surprisingly prolonged occupation. FOOTNOTES [1]We have not been able to determine exactly how many and what size shovel test pits Marian White put into the site. [2] There are records of miscellaneously shaped test pits which may include shovel test pits (above) which together equal 78 square feet. [3] Found on a surface survey 1/2 mile west of Martin site at the south end of Pond No. 3 (Holoway 1964) Statin Island C-14 dates 7410-5310 B.C. (Ritchie and Funk 1973). [4] Laurentian tradition Vergennes Phase Ritchie K-1 site Vermont est. dates 3500-2500 B.C.(Ritchie 1980); McCulley No.1 site on Charlotte Creek in east-central N.Y. 3780 B.C.+/- 110; Shafer site Breakabeen N.Y. 4340 B.C. +/- 100 (Funk and Hartgen 1974). [5] Frontenac Island- 3 C-14 dates 2980-1723 B.C. (Ritchie 1980); O'Neil site Weedsport N.Y. 2050 B.C. +/- 220 and 2010 B.C. +/- 100 (Trautman 1962 p68); Morrisons Island No. 6 Quebec 2750 B.C. +/-150 (Kennedy 1966). [6] Laurentian tradition Vosburg complex 2524 B.C. +/- 300 and 2780 B.C. +/-80 (Ritchie 1961). [7] 3433 B.C. +/- 350 and 2414 B.C. +/- 200 (C-288, Arnold and Libby 1951 p114); 2485 B.C. +/-400 (Crane 1956, p.667); 2575 B.C. +/- 400 (Crane 1956 p668);, 2521 B.C. +/- 300 (Crane and Griffin 1961 p.117); 2451 B.C. 250 (Crane and Grillen 1960 p.38); 2550 B.C. +/- 80 and 2540 B.C. +/- 80 (Crane & Griffin 1962). [8] Laurentian and Frontenac manifestations 2980 B.C. +/- 260 (Arnold and Libby 1951); 2013 B.C. +/- 80 (Y-459) and 1723 B.C. 259 (W-545) (Ritchie 1961). [9] Transitional period Late Archaic Early Woodland 1200-700 B.C. (Ritchie 1961). [10] Transitional Period Archaic/Woodland, Stony Brook 5 C-14 dates 1043-763 B.C. Pelham Bay 876 B.C. +/- 200 (Ritchie 1981). [11] Cowan 1992 Personal Communication. [12] Characteristic point type of early Point Peninsula C-14 date 563 B.C. (Crane and Griffin 1959). [13] 998 B.C. +/- 170 (Libby 1951) Oberlander No. 2 site, Brewerton, Oswego County N.Y. [14] 905 +/- 250 (Crane 1956 p.668). [15] Middle and late Point Peninsula complexes and the characteristic type of the Owasco all stages of development (Ritchie 1961). [16] "the distinctive Iroquoian form" (Ritchie 1961 p.8); Morrisons Island No. 6 Quebec 2750 B.C. +/-150 (Kennedy 1966). [17]This process of association and tabulation leads to two implications of which the reader should be aware. First, if one adds the percentages of the "artifact count of each type of artifact associated to that phase expressed as a percentage of the total number" in all six phases, it will not necessarily equal 100 percent because there were some artifacts which were not found in association with datable objects. Second, in some cases if one adds the percentages of the "artifact count of each type of artifact associated to that phase expressed as a percentage of the total number" in all six phases, the sum may be greater than 100%. This is the result of the problem which arises when artifacts are associated with datable objects from more than one phase. After considerable debate on the methodology the authors decided to count the same artifacts for each phase to which the artifacts could be associated which may result in multiple counting. Alternative strategies could have been splitting the artifact between the phases according to an arbitrary formula such as 50%, 50%; 40%, 60%, etc. [18]The authors are aware that there are some inherent problems with this technique. Namely, if there are one association unit containing two types of pottery which are dated to the different periods, both are counted for each period. Similar problems would exist if the association unit had pottery from three different time periods. The authors tried other means of isolating phases by other combinatorial criteria. For example, a 6 phase system was used. In this system the categories were Vinette 1 for phase a -Early Woodland; Vinette 1 and Point Peninsula for phase b - Early/Middle Woodland; Point Peninsula for phase c- Middle Woodland; Point Peninsula and Owasco Pre-Iroquoian for phase d- Middle/Late Woodland; Owasco Pre-Iroquoian for phase e- Late Woodland; and all three in one association unit, Vinette 1, Point Peninsula, and Owasco- Pre-Iroquoian for phase f-an anomalous Woodland category. (There was only one phase f unit). The rationale for the using the system which was used was based upon its comparability to the point-projectile chronological association scheme. [19] The beaver incisor is not being considered an aberrant pattern since it has only one item in the category. [20] Unifaces, blades, choppers, burins, hammerstones, groundstone and polished stone only occur in the Middle Archaic. Ceramics which should not date to these periods. However, those ceramics which occur follow the standard pattern but should be considered to be an erroneous result of the strict methodology and should be discounted. [21] Lake Erie fisheries have exhibited a high productivity throughout their history. The highest non-stocked production known is over 70 million pounds in 1889 and 1900 or 3.5 tons per square mile. Lake Ontario fisheries produced about .5 tons per square mile. Using volume and outflow figures approximately 106 fish per second should be passing by Grand Island (Zubrow 1980). [22]These figures represent approximately 50% of the faunal material recovered in the three excavating seasons. Completion of the analysis was halted prematurely by budgetary constraints. Analysis courtesy of Patricia Ellis and Howard Savage, University of Toronto Department of Anthropology. REFERENCES CITED Allen, Kathleen 1988 "Ceramic Style and Social Continuity in an Iroquois Tribe" Doctoral Dissertation SUNY Buffalo. Arnold, J.R. & Libby, W.F. 1951 Radiocarbon dates. Science, Vol. 113, No.2927, pp.111- 120 Lancaster. Balson, Jan and Linda Lubas 1978 "Trade on Grand Island" SUNY Buffalo. Birnie,R. and Nagel,B. 1989 "Stage III Cultural Resource Investigations For a Proposed Power Line Location at The Anderson and Locust Point Sites Selkirk Shores State Park Town of Richland, Oswego County, New York" Rochester Museum and Science Center, 657 East Avenue, Rochester, NY 14607. Buerger, Paul T. 1992 "Early Man on the Niagara; The Martin Site On Grand Island, New York" Masters Thesis SUNY Buffalo. Crane, H.R. 1956 University of Michigan radiocarbon dates I. Science, Vol.124, No. 3224, pp. 664-672. Lancaster. Crane, H.R. & Griffin, J.B. 1959 University of Michigan radiocarbon dates IV. American Journal of Science Radiocarbon Supplement, Radiocarbon Vol. I. New Haven. Crane, H.R. & Griffin,J.B. 1960 University of Michigan radiocarbon dates V. American Journal of Science Radiocarbon Supplement, Radiocarbon Vol. 2, pp. 31-48. New Haven. Crane, H.R. & Griffin, J.B. 1961 University of Michigan radiocarbon dates VI. American Journal of Science Radiocarbon Supplement, Radiocarbon, Vol. 3, pp. 105-25 New Haven. Crane, H.R. & Griffin, J.B. 1962 University of Michigan radiocarbon dates VII. American Journal of Science, Radiocarbon, Vol.4, pp. 183-203. New Haven. Ellis, Chris J., Ian T. Kenyon, and Michael W. Spence 1990 "The Archaic" In The Archaeology of Southern Ontario to A.D. 1650, edited by Chris J. Ellis and Neal Ferris, pp. 65-124. Occasional Publication of the London Chapter, OAS 5. Ellis, Patricia J. 1985 "The Martin Site Faunal Report" Department of Anthropology, University of Toronto. Fie, Shannon et.al. 1990 "Encrustation in Iroquois Ceramic Vessels and Food Resource Areas" in Organic Contents of Ancient Vessels: Materials Analysis and Archaeological Investigation edited by William Biers and Patrick McGovern. Museum Applied Science Center for Archaeology. University Museum of Archaeology and Anthropology, University of Pennsylvania vol. 7 pp. 11-24. Funk, Robert E. 1976 "Recent Contributions to Hudson Valley Prehistory" New York State Museum and Science Service. Granger, Joseph 1974 "Riverhaven II Site- Grand Island New York" Doctoral Dissertation SUNY Buffalo. Hayes, Charles F. III ed. 1980 "Proceedings of the 1979 Iroquois Pottery Conference" Research Records No. 13 Rochester Museum and Science Center, Rochester NY. Holloway, Joyce E. 1964 "Aboriginal Occupations of Beaver Island State Park" Masters Thesis SUNY Buffalo. Hunt, Eleazer and White, Marian 1986 "Researching Settlement Patterns Of The Niagara Frontier" North American Archaeologist Vol. 7. Kennedy, Clyde C. 1966 Preliminary report on the Morrison's Island-6 site. National Museum of Canada, Bulletin No. 206, pp. 100- 124. Ottawa. Mellody, Margaret and Kathryn Johnston 1978 "Seed and Nut Identification at the Martin and Burnt Ship Sites" SUNY Buffalo. Noonan, Karen K. 1966 "An Analysis of the Cultural Sequence at the Martin Site: A Transitional Middle/Late Woodland Site" Masters Thesis SUNY Buffalo. Ritchie, William A. 1961 "A Typology and Nomenclature for New York Projectile Points" (Revised 1971) New York State Museum and Science Service Bulletin No.384. Ritchie, William A. 1980 The Archaeology of New York State Rev.Ed. Harbor Hill Books Harrison N.Y. Ritchie, William A. and Funk, Robert E. 1973 Aboriginal settlement patterns in the Northeast. Albany: University of the State of New York, State Education Dept. Sadow, David J. 1980 "The Martin Site's Stone Projectile Points: A Typology and Cultural Chronology" Masters Thesis SUNY Buffalo. Schiepatti, Frank J. 1984 "Intrasite Interpolation Of in Archaeology" Doctoral Dissertation SUNY Buffalo. Trautman, Milton A. 1963 Isotopes. Inc. Radiocarbon measurements III. American Journal of Science, Radiocarbon, Vol. 5, Yale University, pp. 62-79. New Haven. White, Marian E. 1964 "1963 Excavations On Grand Island's East River" Science on the March Magazine of the Buffalo Museum of Science Vol.44 No.3 Buffalo NY. Zubrow, Ezra B. W. 1980 "Problems in Insular Prehistoric Demography" American Anthropological Meetings, Washington D. C. ================================================================= LIST OF TABLES TABLE I Numbers of Excavation Units by Year Excavated TABLE II Martin Site Radiocarbon Dates TABLE III Artifact Tabulation TABLE IV Projectile Point Type Classification TABLE V Rimsherd Type Classification TABLE VI Artifact Associations/Projectile Point dated Cultural Phase VI A Archaic VI B Woodland TABLE VII Artifact Associations/Pottery dated Cultural Phase TABLE VIII Unit Profile 70N 20E TABLE IX Faunal Analysis TABLE X Place of Origin of Flint Tool Material by Cultural Phase Table I Numbers of Excavation Units by Year Excavated Type of Number of Number of Number of Accumu- Excava- Units Units Units lated tion excavated excavated excavated Excava- in 1963 in 1978 in 1979 tion area White Zubrow Zubrow by cat Shovel [1] 217 0 217 test pits sq. 1'x 1' Krig 0 42 15 128.25 units sq.ft. 1.5'x1.5' 5'x5' 84 10 14 270 excavation sq.ft. units 2.5'x5' 0 1 8 112.5 excavation sq.ft. units miscel- [2] 0 0 78 laeous sq.ft. odd shaped units Total 84 270 37 3235.75 Table II Martin Site Radiocarbon Dates 1. Feature 51 2500 BP +/- 120 550 BC +/- 120 2. Feature 9 1630 BP +/- 65 320 AD +/- 65 70 N 20 E 3. 70 N 20 E 1425 BP +/- 140 525 AD +/- 140 4. 70 N 40 W 1325 BP +/- 135 625 AD +/- 135 5. Feature 24 490 BP +/- 70 1460 AD +/- 70 70 N 20 E Table III Artifact Tabulation I. Lithics III. Ecofacts Flint Chunks.........2,712 Bone Fragment.......2813 Flint Cores............241 Fish Vertebra........1 Decortification Flakes...8 Antler...............3 Flint Flakes.......107,002 Beaver Incisor.......4 Retouched Flakes.....3,100 Clam Shell.........128 Unifaces................33 Drilled Shell Bead...1 Bifaces................214 Seed................10 Projectile Points......456 Acorn Shell..........7 Scraper................257 Charcoal...........588 Blades.................161 Soil Sample.........31 Drills..................10 Teeth................1 Choppers................11 Oyster Shell........18 Burins...................3 Bone Drill...........1 Knives..................24 Bone Whistle.........1 Hammerstones............24 Bone Tool............2 Ground Stone Gouge.......1 Snail Shell..........1 Net Sinkers.............53 Nut..................1 Sedimentary Rocks......247 Walnut Shell.........1 Ground Stones...........24 Ground Slate.............3 IV. Metalic Objects Pitted Stones............4 Metal Latch.........36 Polished Stones..........9 Historic Button......4 Quartz..................31 Lead Musket Ball.....3 Quartzite...............59 Nail...............478 Rocks....................8 Pewter Button........1 Fire Cracked Rocks.......4 Iron Projectile Pt. .1 Coal.....................5 Odd Copper Items.....3 Ochre Pieces............18 Copper Coin..........1 Gorget...................1 Wire.................3 Anvil Stone..............1 1817 Dime Cuff Link..1 II.Ceramics V. Miscellaneous Potsherds...........18,879 Shot Gun Shell.......3 Rimsherds..............349 Plastic..............3 Plain Pottery..........162 Slag.................1 Incised Pottery.........46 Roof Shingle.........2 Decorated Pottery.......74 Historic Pottery.......151 Historic Ceramic.......240 Pipe Fragment...........15 Brick..................107 Glass..................378 Bead.....................4 Kaolin Pipe Stem........29 Table IV Projectile Point Type Classification Time Sequence Point Type Estimated Age Frequency (Sadow 1980) Early Archaic Kanawha Stemmed ca. 7410-5310 B.C.[3] 1 Late Middle Archaic Otter Creek ca.4000-3570 B.C.[4] 5 Late Archaic and Transitional 41 Brewerton Side-Notched 21 Brewerton Corner-Notched ca. 2980-1723 B.C.[5] 1 Brewerton Eared-Triangle Normanskill ca. 2800-2500 B.C.[6] 1 Lamoka ca. 2500-2000 B.C.[7] 3 Genesee ca. 2500-2000 B.C.[8] 8 Susquehanna Broad ca. 1043- 763 B.C.[9] 2 Perkiomen Broad ca. 1043- 763 B.C.[10] 4 Dry Brook Fishtail ca. 1600 B.C.[11] 1 Early Woodland 16 Meadowood ca. 800-400 B.C.[12] 15 Fulton Turkey-Tail ca. 600 B.C.[13] 1 Middle to Late Woodland 24 Jack's Reef Corner-notched ca. 200-700 A.D.[14] 5 Levanna ca. 700-1350 A.D.[15] 19 Iroquoian Late Woodland Madison ca. 1200-1600 A.D.[16] 43 Archaic Points 36% 47 Woodland Points 64% 83 Total 130 Table V Rimsherd Type Classification Undecorated 28 Decorated 139 Early Woodland Vinette Type I 22% 31 Middle Woodland Point Peninsula 71% 99 Late Woodland Owasco pre-Iroquoian 6% 9 Table VI Artifact Associations/Projectile Point dated Cultural Phase Cultural Phase Artifacts Late Middle Archaic Late Archaic + Trans. Lithics Tnum %T AwtG Tnum %T AwtG Flakes 682 1 1.4 17068 17 1.5 Cores 2 2 25.0 Ret.Flakes 16 1 3.0 321 11 4.7 Chunks 85 2 18.3 Unifaces 3 1 23.0 Bifaces 6 1 3.4 34 18 6.3 Points 4 1 4.4 91 21 5.4 Scrapers 3 1 6.2 40 18 8.5 Blades 4 2 4.3 Drills 7 30 8.1 Choppers 1 9 38.0 Burin 1 33 15.0 Knives 3 16 16.3 Hammerstones 2 8 360.0 Decort.Fla. Net Sinkers 11 21 380.5 Sed.Rock 119 47 10.1 Ground Stones 9 37 14.1 Ground Slate Polished Stones 3 33 144.7 Quartzite Hematite Burnt Flint Flint Needle Anvilstone Gorget Ceramics Potsherds 86 1.9 1.9 4607 24 5.8 Rimsherds 167 49 3.0 Hist.Pot. 48 33 2.3 Hist.Cer. 1 0.01 2.8 27 11 2.6 Pipe Frag. 1 6 2.8 1 6 4.0 Brick 14 9 3.9 Glass 5 10 1.8 Kaolin Stem 18 46 2.2 Ecofacts Bone Frag. 433 17 0.6 Beaver Incisor Clam Shell 64 41 0.8 Charcoal 339 49 3.6 Soil Sample Ochre Pieces 1 6 145.1 2 12 48.7 Oyster Shell Nut Bone Tool Walnut Shell Misc. Latch+Screw 1 3 5.0 Nails 46 9 4.1 Table VI (continued) Artifact Associations/Projectile Point dated Cultural Phase Cultural Phase Early Woodland Middle Woodland Late Woodland Artifacts Tnum %T AwtG Tnum %T AwtG Tnum %T AwtG Lithics Flakes 5747 5 1.7 14986 14 1.6 14225 13 1.5 Cores 13 5 32.0 65 27 40.0 7 3 23.0 Ret. Flakes 134 4 5.0 273 9 6.0 409 13 5.0 Chunks 50 2 4.9 421 15 2.1 58 2 2.6 Unifaces 3 9 8.6 2 6 8.2 Bifaces 14 6 8.3 29 14 8.4 28 14 8.3 Points 35 8 5.2 72 16 5.7 48 11 5.6 Scrapers 23 9 8.2 37 14 10.2 31 12 10.6 Blades 3 2 8.0 7 4 3.4 21 13 2.2 Drills 4 16 20.0 7 46 17.2 4 26 18.4 Choppers 1 9 30.0 2 18 30.0 1 9 30.0 Burins Knives 2 8 7.5 4 16 18.1 1 4 18.4 Hammerstones 1 4 131.0 3 13 290.0 Decort.Fla. 2 25 36.0 Net Sinkers 5 9 263.0 7 13 190.2 1 2 197.6 Sed.Rock 1 4 41.0 3 1 39.6 114 47 10.0 Ground Stones 6 25 72.6 5 20 68.9 Polished Stones 3 33 182.0 1 11 166.0 Quartzite 6 10 7.0 3 5 5.0 Hematite 1 2 3.0 1 2 0.1 21 36 5.2 Burnt Flint 1 50 1.0 Flint Needle 1 50 41.0 Anvilstone Gorget Ceramics Potsherds 1109 6 3.3 2851 15 3.1 3306 18 2.6 Rimsherds 11 3 11.2 70 20 8.7 48 13 7.9 Hist.Pot. 2 1 0.8 20 13 2.1 30 20 1.8 Hist.Cer. 7 3 4.9 20 8 2.6 8 3 3.1 Pipe Frag. 2 11 1.5 15 6 3.3 13 5 2.5 Brick 1 0.9 0.5 1 1 5.0 14 13 20.8 Glass 6 2 5.5 2 0.5 8.0 14 4 4.7 Kaolin Stem 1 3 5.0 11 25 26.0 9 20 2.3 Ecofacts Bone Frag. 276 10 0.6 398 14 0.5 245 9 0.8 Beaver Incisor Clam Shell 29 23 14.0 64 50 0.6 Charcoal 56 10 1.1 57 10 1.3 107 18 1.2 Soil Sample Ochre Pieces 9 2 6.0 5 28 14.5 2 11 1.5 Oyster Shell Nut Bone Tool Walnut Shell Misc. Latch+Screw 1 3 5.0 Nails 5 1 3.6 66 14 4.0 13 3 3.8 Table VII Artifact Associations/Pottery dated Cultural Phase Phase Early Woodland Middle Woodland Late Woodland Artifacts Tnum %T AwtG Tnum %T AwtG Tnum %T AwtG Lithics Flakes 13337 12 1.6 25232 24 1.7 5943 6 1.7 Cores 4 2 21.9 29 12 24.9 1 0.4 18.7 Ret. Flakes 222 7 6.6 1029 33 2.8 59 2 4.8 Chunks 127 5 17.0 1193 44 2.7 13 0.4 21.9 Unifaces 3 9 22.8 19 56 9.8 5 15 12.6 Bifaces 26 12 10.3 62 29 10.3 18 8 6.7 Points 92 20 3.5 150 33 4.1 25 5 5.3 Scrapers 30 11 14.3 73 28 8.7 15 6 17.6 Blades 2 1 2.8 40 25 2.2 Drills 10 40 11.3 13 52 17.5 4 16 15.7 Choppers 1 9 30.0 5 40 14.8 2 18 34.0 Burins 1 33 15.0 1 33 15.0 1 33 15.0 Knives 4 16 16.9 3 13 20.6 3 13 12.6 Hammerstones 2 8 273.4 7 29 231.4 Decort. Fl. 2 25 10.2 Net Sinkers 10 19 364.4 28 52 168.2 2 4 360.6 Sed. Rock 4 2 32.0 9 4 90.2 1 0.4 40.9 Ground Stones 8 33 3.8 10 41 43.0 8 33 3.8 Ground Slates 2 66 14.7 1 33 14.0 1 33 14.0 Polished Stone 5 56 571.8 4 44 192.4 4 44 192.4 Quartzite 12 20 28.2 12 20 28.2 11 18 30.5 Hematite 1 50 1.0 Burnt Flint 2 29 17.4 Anvil Stone 1 100 1044.0 Gorget 1 100 81.0 Ceramics Potsherds 3985 21 4.0 8232 44 3.7 1144 6 6.1 Rimsherds 235 67 5.1 289 82 5.7 75 21 4.6 Hist.Pot. 22 15 14.3 53 35 5.7 4 3 2.8 Hist.Cer. 23 10 2.1 38 16 2.3 8 3 2.4 Pipe Frag. 2 13 9.4 4 27 6.1 Bricks 1 1 5.4 4 4 13.0 Glass 2 1 0.9 71 19 2.2 1 0.5 4.0 Kaolin Stem 2 50 1.4 1 25 4.9 1 25 4.1 Ecofacts Bone Frag. 324 12 0.6 675 25 0.6 189 7 0.4 Beaver Inc. 1 100 4.4 1 100 4.4 Clam Shell 60 47 0.8 58 45 0.9 19 15 0.7 Charcoal 85 14 0.6 105 18 1.3 83 14 0.6 Soil Sample 6 19 148.8 12 39 151.8 5 16 1.5 Ochre pieces 3 16 3.8 Bone Tool 1 50 5.8 1 50 5.8 Misc. Nail 67 14 2.7 115 24 5.6 4 1 4.1 Table VIII Unit Profile 70N 20E Stratigraphy Surface Level A (top 6") Level B 81 Flakes 195 Flakes 6 Cores 1 Ret. Flake 5 Ret. Flak 5 Chunks 6 Chunks 1 Projectile point 1 Point (Levanna) (Levanna) 1 Biface 2 Decortification Fl. 49 Potsherds 187 Potsherds Rimsherd 1 Rimsherd 4 Rimsherds (Point Peninsula) (Point Peninsula) (Pt.Peninsula) 1 Acorn Shell Features Two Hearths Three Storage or Trash Pits Radiocarbon Dates Feature 9- 320 AD +\-65, 525 AD +\-140 Feature 24- 1460 AD +\-70, Table IX Faunal Analysis Mammels White-tailed Deer.................. 1235 Raccoon.............................121 Grey Squirrel........................61 Muskrat..............................82 Chipmunk.............................36 Beaver...............................25 Black Bear...........................20 Dog or Wolf..........................19 American Elk.........................17 River Otter..........................12 Eastern Moose.........................4 Squirrels, Woodchucks, Porcupines, Rodents,Rabbits Fishes Walleys............................1088 Freshwater Drum.....................231 Channel Catfish.....................134 Lake Sturgeon........................26 White Bass...........................10 Largemouth Bass.......................7 Smallmouth Bass.......................7 Muskellunge...........................2 Unidentified fish vertebrae........1223 Unidentified bivalves................26 Large amounts of fish scales Birds Mergansers...........................93 Greater Scaups.......................19 Canvasbacks..........................16 Blue-winged Teals....................15 Ring-necked Ducks....................12 Canada Geese..........................9 Ducks unidentifiable as to species...79 Wild Turkeys.........................17 Passenger Pigeon......................1 Ruffled Grouse........................2 Horned Grebes........................20 These figures represent approximately 50% of the faunal material recovered in the three excavating seasons. Completion of the analysis was halted prematurely by budgetary constraints. Analysis courtesy of Patricia Ellis and Howard Savage, University of Toronto Department of Anthropology. (Ellis 1985) Table X Place of Origin of Flint Tool Material by Cultural Phase Late Archaic Normanskill Point (10N 0W) Diver's Lake Onondaga-W.N.Y. Normanskill Point (10N 5W) Diver's Lake Onondaga-W.N.Y. Blade Tip Leroy Onondaga-Cent.N.Y. Point Tip (conc. 5) Normanskill Shale-East.N.Y. Utilized Flake Vera Cruz Shale-Pennsylvania Transitional Orient Fish-Tail Point Diver's Lake Onondaga-W.N.Y Early Woodland Fulton Turkey-Tail Point Normanskill Shale-East. N.Y. Meadowood Point Diver's Lake Onondaga-W.N.Y Jack's Reef Point Lockport Flint-Western N.Y. Middle Woodland 2 Blades Flint Ridge Chalcedony-Ohio Unspecified Woodland 2 Flakes Flint Ridge Chalcedony-Ohio (Balson and Lubas 1978)