CHAPTER 5

5.0 The Connecticut Late Archaic and Terminal Archaic:

A Mid-Range Perspective

For a long time, Connecticut amateur archaeologists have recognized the presence of some evidence of the Laurentian. Private collections, in many instances, contain an inordinate number of Brewerton and Vosburg projectile points. Eastern Connecticut collections such as those of Basto, Sterry, Golet, Dill, Greiner, and the Archaeological Society of Southeastern Connecticut (ASSEC) show surprising uniformity in Brewerton eared triangle and side-notched point styles. These specimens are predominantly made of quartzites. However, there are some cobble quartz and occasional chert specimens in collections. Presumably associated tools are drills, scrapers, slate ulus, gouges, basalt axes, adzes, and shale bannerstones. Plummets are not frequently represented, but three gneissic examples from East Haddam's Salmon River site have recently been brought to my attention as well as one from Baltic.

Based upon our growing knowledge of the Late Archaic, a brief review of the older literature from the Archaeological Society of Connecticut Bulletin yields site reports that documented Late Archaic Lake Forest assemblages. This, together with recent publications, produces a convincing argument for the significance of a Connecticut Lake Forest tradition during the Late Archaic.

In the lower Quinnipiac River valley, Sargent (1952) reported projectile points very similar to Brewerton eared triangles from Zone One, a pre-ceramic horizon of Grannis Island. One such specimen was excavated from a discernible level that set itself apart from other levels by the high frequency of basalt debitage.

Glynn (1953), from analysis at the Pilots Point site, as well as the Manstan site in Westbrook, reported Brewerton-like points in the array of other collected specimens. At the Manstan site, a large stone "ring" feature, partially eroded, was found below the high water mark. Over three hundred points were found in and immediately around it. Artifacts usually associated with Lake Forest as well as Mast Forest were collected along with knives, scrapers, drills, choppers, adzes, gouges, and a notched bannerstone.

Coffin (1963) reported some "classic" Brewerton points from his excavated collection from Milford. Keener (1965), from excavations made in the late fifties at the Phillips site in Glastonbury, documented several Lake Forest forms coming from a deep level that he designated Level #5. MacCurdy (1962) described a rockshelter in New Haven that he had salvaged toward the beginning of the century. In it were excellent examples of Lake Forest equipment including a bannerstone, ulu, and adze. Projectile points appeared to be representative of both the Lake Forest and Mast Forest cultural traditions.

Basto (1937) and Praus (1945) reported on the South Woodstock site in northeastern Connecticut. The collection, which is now curated in the Peabody Museum at Yale University, indicates a sizeable Lake Forest component. It is also important to mention that this site witnesses a major technological concentration on the local Plainfield formation quartzite.

The more recent studies and excavations that have employed better techniques and have been aided by multi-disciplinary analyses illustrate a refinement in our ability to recognize Mast and Lake Forest information in Connecticut. Studies by McBride (1984a&b; Pfeiffer 1984) in eastern and southeastern Connecticut have documented a Lake Forest presence that Lavin (1985, 1988) characterizes as an "indigenous resident" population. McBride (1984b:243) downplays the significance of the Lake Forest component in eastern Connecticut. While suggesting that the cultural manifestation be assigned to the Golet phase, he also points out that settlement pattern was very dispersed and that there was no evidence for seasonal or base camps. He envisions a pattern that, "reflects the exploitation by mobile groups of a wide range of micro-environments and environmental zones." This is a stand initially posed by McBride and Dewar (1981). Finally, McBride states that "the (Golet phase) artifacts are few in number and rarely significant at any site." The (McBride 1984b) assessment is based on the interpretation of only a few sites. In his analysis of the Lower Connecticut Valley, he references the Woodchuck Knoll site where a small "Golet phase" component was overlain by a number of Mast Forest components (McBride 1978).

McBride does not consider the sites analyzed herein. He is therefore left with two sites, Woodchuck Knoll and the very small site Coudert Ledge. In his dissertation he discussed Ashland Lake dated 4280 + 120 BP [Beta 5793] and Salmon Cove dated 4180 + 90 [QC-1157], but could not identify the actual sizes and composition of the components.

A surface survey of Pachaug Lake in northeastern Connecticut has yielded similar findings to the Bashan Lake (Pfeiffer 1983). Distinct loci directly relating to the Lake Forest tradition have been identified along the shores of the dammed course of the Pachaug River. The discovery was made during winter lake draw-down in 1988. Large hearths not unlike the ones suggested by Glynn (1953) or Pfeiffer (1983), Brewerton eared triangular points, knives, ulus, drills, scrapers, pestles, mullers, hammerstones, and broken bannerstones were found at these loci. The reduction of Plainfield formation quartzite quarry blocks into bifacial preforms appears to be a major activity at these sites. From one site over four hundred broken in-process preforms have been found in workshops. A strong case can be made for the site's integrity because broken artifacts can be rearticulated from fragments littering the workshop surface. The projectile points found in and surrounding such workshops have been identified elsewhere in eastern Connecticut and are now considered diagnostic Lake Forest types.

Projectile point analysis of both surface and excavated assemblages from eastern Connecticut indicates that Brewerton eared triangles are most prevalent, but some specimens grade stylistically toward Vosburg. There is a higher incidence of Vosburg points as analysis proceeds westward toward central Connecticut. In the Quinnipiac River valley, for example, many sites have yielded both Vosburg and Brewerton forms; however, here the greater proportion are Vosburgs (Maurice Wilson, 1983 personal communication; Lavin and Russell 1985). Quartzite materials are not common in more central zones of Connecticut; instead, red felsite predominates (Lavin and Russell 1985).

In central southern Connecticut, Lake Forest components are reported at the Fastener site (Lavin and Salwen 1983). Lavin suggests that there is stratigraphic evidence for the contemporaneity of these with narrow stemmed points of the Mast Forest. At the Burwell-Karako site the "Laurentian" is documented by thirty-one Vosburg and Brewerton points. Lithic materials are suggested to be relatively local in origin (Lavin and Russell 1985:50-51). As in coastal central Connecticut, local quartz is also evidenced at the Fastener and Burwell-Karako sites. Lavin and Russell (1985:52) state that the local level of lithic acquisition may help support the theory that there were well established residential "Laurentian" groups in central Connecticut.

In southwestern Connecticut Wiegand (1983) studied twenty rockshelters to determine: (1) relationship of topography to the frequency of rockshelters (2) the determining factors of rockshelter use in prehistory and (3) site function. The reports of the project suggest a significant Laurentian presence at these shelters. Of the thirteen shelters that witnessed habitation in prehistory, six showed Laurentian artifacts. In this area, apparently the Vosburg type predominated (Wiegand 1983:162).

According to Swigart (1974) there are only a few Vosburg and Brewerton projectile points in western Connecticut, which suggests a less than significant Laurentian presence. However, Thompson (1969:14-15) reported a Vosburg component, dated to 4340 + 120 BP [Y-1664] for the Binette site, a rockshelter in Naugatuck in western Connecticut. Of the specimens known in the Housatonic and western Connecticut drainage areas, quartzite is rarely used as a raw material in the production of Vosburg and Brewerton projectile points. Instead, for western Connecticut, New York cherts are far more common.

Thompson (1988:40) reports several occupations from the Sugar Loaf Hill rockshelter site in the northwestern corner of the state. Stratigraphically overlying a Middle Archaic component was a Vosburg projectile point. He suggests that this cultural manifestation was abruptly replaced by Mast Forest forms. While there are no dates to calibrate such observations, the progression of components from Middle Archaic through Late Archaic including a Lake Forest unit may be significant.

Swigart (1987) found a poorly stratified situation in the Woodruff Rockshelter in northwestern Connecticut. He attributes this to the incidence of frequent rockfalls. Toward the bottom of the excavation were Otter Creek, Brewerton, small stemmed, and triangular projectile points along with a section of bannerstone.

It is clear from the literature that McBride's (1984b) assessment of the Lake Forest presence in Connecticut, while trying to accommodate Dincauze (1975), under-represents the archaeological record. There is sufficient evidence from many sites and site types to support a southern New England expression of the Lake Forest. The individual Lake Forest back country sites may appear insignificant as McBride (1984) suggests from his site specific perspective. But taken together, evidence suggests that a strong case can be made for there being a meaningful pattern. McBride (1984b:242) leaves the door open on this issue by suggesting the possible existence of larger sites with Lake Forest components.

What can be interpreted from the recent literature is that settlement pattern may have been complex. There is evidence that there are different site sizes as well as corresponding range of activities. The back country sites could have likely functioned as single activity, task specific loci. The larger sites that have many activities represented in their archaeological deposits probably were base camps. While the site types are suggested in the literature, the specific relationship of these sites to one another is poorly demonstrated.

Figure 5.0.1

Connecticut Late Archaic Lake Forest Tradition Dates and Calibrations
 

Site	Laboratory  Lab  #  Date	Cal BP	Cal BC	Max/Min BP	Max/Min BC
Ashlnd Lk                      Beta	5792    4280+120	4860	2911	4979-4650	3030-2701
Sal Cove                         QC	1157     4180+90	4824, 4751, 4727    4664, 4655	2875, 2802, 2778    2715, 2706	4895-4564	2910-2615
Binette                    Y	1664 4340+120	4871	2922	5202-4735	3253-2786

5.1 The Connecticut Late Archaic Mast Forest Tradition

There is general agreement among Connecticut archaeologists that during the Late Archaic's final centuries, there was a cultural adaptation to the Mast Forest ecological setting. There is not a consensus with respect to the terminology of this archaeological phenomenon, but all accept the validity of the numerous corresponding features. This cultural manifestation is represented within many multi-component sites over the entire state. There is also clear evidence from the eastern portion of the state that this culture, with its diagnostic narrow stemmed projectile points, persisted through time into the Terminal Archaic period and that the points, at least probably continued into the early Woodland period (Tuck 1978; Funk 1984; McBride 1984a&b; Lavin and Russell 1985; Lavin 1988; Hoffman 1985, 1988; Pagoulatos 1988). Because of the persistence of the point type and the quartz cobble technology, there have doubtlessly been errors as to assignment of period for some sites. This may account for some inflation of population estimates during the Late Archaic and the reverse for the Terminal Archaic and Early Woodland periods (Lavin 1988:110; Hoffman 1985:64).

The characteristics of the narrow stemmed point in Connecticut are, (1) blade with straight or slightly convex edge, (2) shoulders of very weak to moderate form merging into a long parallel-sided to slightly contracting stem, and (3) a straight or slightly convex base, often thick and unfinished. Preferred materials are local quartz, local quartzite, local cherts, and sometimes local slate. An important feature here is the "local flavor" of specimens (Ritchie 1961, Snow 1980, Fogelman 1988).

A second diagnostic point type of the Mast Forest tradition in Connecticut is the "Squibnocket" triangle. Its characteristics are a convex or straight-sided point with a straight or slightly convex base. Such specimens are fairly well made and usually small, 1.5 cm up to 3 cm. Their triangular forms are sometimes equilateral, sometimes isosceles (Ritchie 1961).

The Mast Forest Archaic has been dated in Connecticut within the Connecticut Valley at Hubbard Brook in Middletown to 4280 + 120 BP [Beta 5793]. Farther upriver in Glastonbury, Long Knoll was dated to 3995 + 100 BP [QC-976] (McBride 1984a:60). In the Housatonic Valley at site 6 LF 2 dates of 4465 + 80 BP [GX2489] and 4355 + 185 BP [GX2696] were reported by Swigart (1974). A date of 3800 BP was determined at site 6 LF 1 (Swigart 1974). McBride (1978) reports two dates of 3570 BP and 3690 BP from the Woodchuck Knoll site. Both dates are associated with narrow stemmed projectile point levels. However, as McBride points out, the more recent date was generated from a sample taken beneath the level that generated the older date. For this Late Archaic period, dates across the state range from 4500 to 3700 BP.

McBride (1984a:66-69) subdivides the Mast Forest Archaic into two phases. The Vibert phase is characterized by triangular formed points, that are potentially slightly older and found in strata that are beneath narrow stemmed points. The Tinkham phase is characterized by the narrow stemmed projectile points and is more recent.

By reviewing some of the more recent literature pertaining to the Mast Forest tradition, there is a strong indication that such groups manufactured projectile points from local materials (Lavin 1984, 1988; Lavin and Russell 1985; McBride 1984a&b; Wiegard 1983; Feder 1981). Associated with the usage of local material is a strong reliance upon a cobble industry. In such a technology, tools are produced from cobbles rather than quarry blocks. However, in areas where quarry sources predominate, especially back from the coastline, river and streambeds, study of debitage and tools shows the use of quarried material but rarely to the complete exclusion of cobbles (Lavin, McBride personal communications N.D.).

In the Quinnipiac valley Lavin and Russell (1985:52-54) at the Burwell-Karako site, defined two specific types within the Burwell complex. The Burwell type was defined as points with equilateral triangular blades with short but wide stems that are often tanged or eared. These researchers suggested that this type stylistically intergrades with Brewerton eared type points and may indicate some level of interaction between groups. In some units of this excavation, Burwell points appeared to be stratigraphically between Brewerton and narrow stemmed points, therefore implying a temporal relationship as well. This type is probably comparable to points described by McBride (1984a&b) for the Vibert phase.

Lavin and Russell (1985:54) also defined a Karako type point which usually, with respect to stratigraphy, overlay Burwell points at the site. The Karako "type" was described as a flat stemmed point with a wide equilateral triangular blade and sharp distinctive shoulders. These authors suggested that it could be a variant of the Wading River type or possibly a small quartz Snook Kill point.

Other items reported from such sites are hammerstones, pitted stones, axes, and pestles. However, there has been a paucity of larger ground stone woodworking tools reported in the literature relating to Mast Forest sites. This characteristic was noted at Hubbard Brook (Ingui 1986), Salmon River (Golet personal communication 1981) and other Mast Forest sites excavated and/or collected by amateurs. There is the possibility that such items were previously recovered and thus not reported or that some had gone unrecognized. Yet it is also possible that this pattern is real and that ground stone tools are not common within the Mast Forest assemblage. The technological subsystem is therefore best known for its narrow stemmed points and the use of local materials. Of significant note is the cobble industry.

Mast Forest tradition subsistence was diffuse and opportunistic (McBride 1984a&b). Site location and activities represented within these sites have suggested both special purpose upland camps (Keener 1965) and seasonal camps oriented toward riverine, interior wetlands and lacustrine environments (McBride 1984a&b).

Subsistence evidence at the seasonal camps of Woodchuck Knoll and Haddam Meadows revealed the use of American lotus, goosefoot, walnut, and hickory. Also indicated was that some of these may have been stored (McBride 1978,1984 a&b). Fishing was indicated by weirs and netsinkers at Bashan Lake (Pfeiffer 1983). Freshwater clam, oyster, and quahog suggest the usage of both aquatic and marine resources. Bone within midden and hearth deposits suggest that deer as well as other mammals and birds were being exploited from terrestrial environments (Praus 1942, Coffin 1963, Salwen personal communication 1972 [Shantok cove], Wiegand 1982, McBride 1984a&b, Lavin and Russell 1985, Lavin 1988).

The ideological subsystem as evidenced by burial practice in the Mast Forest Archaic is not well known and thus far has not been specifically defined. Primary flexed burials have been discovered with narrow stemmed points, quartz cobbles in various stages of reduction, and antler flaking tools in several sites excavated over the years (Praus 1942; Coffin 1963). These features, however, were not dated and could not specifically be assigned to the Archaic, Terminal Archaic, or Early Woodland period. Considering the conservatism of various groups with respect to the ideological subsystem (Steward 1955), it would not be surprising if this form of burial custom was representative of the Mast Forest tradition.

Figure 5.1.1 Connecticut Mast Forest Tinkham Phase Dates and Calibrations
 

Site	Laboratory Lab # Date	Cal BP	Cal BC	Max/Min BP	Max/Min BC
Hubbard Bk                          Beta5793	4280+120	4860	2911	4979-4650	3030-2701
Long Knoll        QC976	3995+100	4510,4492,4446	2561,2543,2497	4803-4356	2854-2407
6 LF 2                GX2489	4465+80	5207,5197,5049	3258,3248,3100	5289-4881	3340-2932
6 LF 2                GX2696	4355+185	4964,4953,4874	3015,3004,2925	5289-4653	3340-2704

5.2 Mast Forest Related Occupations during the Terminal Archaic Period in Connecticut

The Terminal Archaic period has been generally considered the "period of the Susquehanna tradition". Yet, it is becoming clear that narrow stemmed projectile points of presumably the Mast Forest occupations persisted from Late Archaic times. According to McBride (1984 a&b), Lavin (1984, 1988), Pfeiffer (1984, 1990), and Pagoulatos (1988), the settlement system appears to take a more upland character during the Terminal Archaic Period. Sites that demonstrate this aspect are Candlewood Hill dated to 3470 + 60 [Beta 3950], Beaver Brook at 3130 + 90 BP [Beta 7071], Cedar Lake at 2910 + 110 BP [Beta 5317], and Binette 2800 + 140 BP [Y-2582] (Swigart 1974, McBride 1984b). There may also be a second site type that corresponds directly to marine shellfish resources. Such sites are along river mouths and ocean fronts. Examples of this, while not in Connecticut, are Hornblower #II (Ritchie 1969b) on Martha's Vineyard and the Two Springs site on Fishers Island (Funk and Pfeiffer 1988:103). These sites document narrow stemmed projectile point and presumably a Mast Forest presence before a Broad Spear complex habitation and again after. Dates from components in coastal sites bracket the Terminal Archaic period. The stratigraphic relationship of Mast Forest components appearing both above and below Broad Spear complex components has been seen at the Burwell-Karako site (Lavin and Russell 1985: 58) Ely's Ferry site, and Long Knoll site (McBride 1984a:60-61) to name a few. This stratigraphy has also been shown to me by an amateur at a rockshelter site along the Quinnebaug in northeastern Connecticut (Greiner personal communication 1985).

Figure 5.2.1 Connecticut Terminal Archaic Sites which have Narrow Stemmed Projectile Points Dates and Calibrations
 

Site	Laboratory Lab # Date	Cal BP	Cal BC	Max/Min BP	Max/Min BC
Cndl Wd Hl              Beta3950	3470+60	3815,3795,3721	1866,1846,1772	3835-3654	1886-1705
Beaver Bk Beta7071	3130+90	3369	1420	3463-3262	1514-1313
Cedar Lake  Beta5317	2910+110	3049	1100	3249-2882	1300-933
Binette   Y-2581	2800+140	2925,2914,2882	976,965,933	3109-2769	1160-820

5.3 Connecticut Terminal Archaic and The River Plain Tradition

Various terms have been applied to the best known cultural manifestation of the Terminal Archaic: Susquehanna tradition and Broad Spear, to name a few. Snow (1980) has suggested that such sites represent cultural developments grafted on to the Mast Forest cultural adaptation of southern New England. In his view, these expressions are characterized by technological traits that had diffused into southern New England and became attached to the burial complex.

Much evidence has recently become available to demonstrate an "other than mortuary context" (Snow 1980) for broad spears (McBride and Dewar 1981, Pfeiffer 1984, Lavin 1984, Pagoulatos 1988). Well dated habitation sites with excellent broad spear associations are coming to light in the Connecticut River valley, and these suggest a complete cultural system. I refer to this Terminal Archaic culture system as a River Plain tradition in accordance with Tuck's (1978) and Snow's (1980) attempt to apply terms which reflect methods of adaptation and the physiographic setting. Such sites are described by McBride (1984 a & b), Lavin (1984, 1988), Lavin and Russell (1985), Pagoulatos (1988), Thompson (1989), and Wilson (personal communication, 1983).

The River Plain tradition of the Terminal Archaic period 3700-2700 BP is identified by several point types which Dincauze (1968, 1972, 1975) and Snow (1980) believe represent temporal phases. Pfeiffer (1984, 1990), McBride (1984b), Lavin and Russell (1985) suggest that the typological variability may not necessarily be completely a consequence of time.

When projectile point types have been used to delineate phases in this Terminal Archaic sequence, three such entities have been described. I am not completely at ease with this scheme because it is built on typological characteristics that purportedly are temporally distinct. The radiometric determinations are not clear in these cases and types have sometimes been defined on the basis of single artifacts in dated pits. The literature, however, indicates that other researchers are more confident in using these distinctions.

The first point type that has been defined as an early representative of this tradition is Snook Kill, dated in New York to approximately 3600 BP. Its counterpart in New England and Connecticut is seen in Dincauze's (1972) formulation of the Atlantic phase. In Connecticut there is a small but even distribution of such material. Sites of the phase are usually open-air and overlook rivers (Pfeiffer 1984, McBride 1984b, Lavin 1988, Pagoulatos 1988). A small discrete locus in the LaRose site in South Windsor had a dramatic assemblage of such points in a habitation context. This site was surface collected during a major construction project. In the midst of such construction I learned about the site and brought a survey team up to test the area. At this time, I noted discrete burial and habitation loci.

Cremation burials may belong to a second Broad Spear phase in Connecticut and have been termed the Watertown phase by Dincauze (1968). Outside the limits of the study area are the dated sites of Schwartz in South Windsor and Toelle Road in Wallingford (Pfeiffer 1984,1990). The Toelle Road site was excavated in the early sixties by untrained collectors and fortunately a more knowledgeable amateur who took me back to the site in 1983. Maurice Wilson and I re-excavated several of the burial pits finding that these were never totally removed. We retrieved several Broad Spear points, charcoal for dating, and acorn, hickory nuts, and chenopodium in the greasy matrix of these partially disturbed features. The charcoal sample was sent to the Smithsonian for age determination and Toelle Road site was dated to 3055 + 120 [SI-5313].

The Schwartz site was excavated in the early seventies by members of the Albert Morgan Chapter of the Archaeological Society of Connecticut and the State Archaeologist. The site was prolific and in comparison to the Griffin site had the similar trait of variability in the assemblage. The excavation was made under severe salvage conditions and the internal structure of the site is poorly known. A charcoal sample was given to me by the State Archaeologist, Douglas Jordan, and the date that was derived by the laboratory for the Schwartz site was 3335 + 90 [GX-10850]. Dated burial sites range from 3535 to 2985 BP (Pfeiffer 1980, 1984, 1990) suggesting a hiatus within the second half of the third millennium BP. In 1983, McBride and I worked together in this study area and combined our resources to identify habitation sites with broad spear points and features. From these sites have come the dates: Site 105-34, 3610 + 70 BP [Beta-7808] Brodeur Point, 2970 + 85 BP [SI-5309]; Site 105-33, 2700 + 60 [Beta-7810]; and Cooper's, 2740 + 90 BP [Beta-5314].

A third phase, characterized by Orient cremation burials (Ritchie 1959) has not been described in Connecticut, but such sites have been reported in Rhode Island and on Long Island. Projectile points of the Orient phase are found throughout Connecticut, mostly along watercourses.

Private and museum collections document a significant River Plain tradition in Connecticut (Lavin 1988, Lavin and Russell 1985). Site locations occur predominantly along watercourses of deep water and slow currents (Pfeiffer 1984, 1990). Pagoulatos (1988) notes the location of such sites as overlooking floodplains at the margins of terraces. There is very little inland or upland evidence of this phase (McBride 1978, 1984a&b). Ritchie (1969a) and Funk and Rippeteau (1977) observed the same pattern in New York.

Reliable local resources were apparently critical to the location of habitation sites. The sites lie just above the river flood plain where floral and faunal resources were readily available (Pfeiffer 1984,1990; Pagoulatos 1988). Although separated from them by a hundred or more meters, cremation burials have been identified in close proximity to habitation sites (Pfeiffer 1984,1990). The Toelle Road site in Wallingford (Wilson, personal communication 1983) and the Carrier site in Glastonbury are both burial sites and have nearby, closely related habitation loci. Pagoulatos (1988) studied the Timothy Stevens site, which was potentially a related habitation for the Carrier burial site. The South Woodstock site (Basto 1937) showed a similar relationship between occupations and burials.

The cremation burial sites are both dramatic and prolific and have given us our first insights into this Terminal Archaic period culture. Because these were the first sites that were identified for this tradition, it is quite possible that archaeologists formed unrealistic expectations of equally grand and prolific habitation sites. This partly accounts for some of the difficulty in initially recognizing the real nature of this culture's habitation sites.

The cremation burial sites of Griffin (Pfeiffer 1980 a & b), Schwartz (Jordan ND), Carrier (Jordan ND), Laurel Beach and Minor (Lavin 1988), Rye Hill (Thompson 1989), Toelle Road (Wilson and Applegate personal communication 1983) and others serve as good examples of what was in the complex. Steatite bowls, large broad bifacial blades, "Susquehanna" type projectile points (Dincauze 1968, Lavin 1988), heavy wood working tools, grinding pestles, sharpening stones and food offerings were found in the burial features. Some artifacts were ritually "killed" (Dincauze 1968, Ritchie 1969a, Pfeiffer 1980a&b); others were placed unburned into pits after cremation. Lithic materials from some sites have been shown to be exotic, coming from hundreds of kilometers miles away (Pfeiffer 1984). Yet at other burial sites, more local sources appear to have been used (Lavin 1988). There may well be some significance to the observation of differential sources of raw material for these sites. When correlated with the age determinations for these same sites, the older cremation burial sites seem to have more exotic materials than the burial sites that occur towards the end of the Terminal Archaic period. More C14 dating and analysis would test the validity of this possible pattern. If in fact this is true, there may be a basis for delineating cultural phases, extraneous to the more traditional typological approach used by Dincauze (1968,1972).

Dry bone cremation is suggested by most of the researchers of these sites, who are probably using Dincauze (1968) as a model. However, there is too little osteological study to back up such statements. A review of the ASC Bulletin and other early twentieth century documents reveals a plethora of River Plain tradition material. Beyond the early work of Basto (1937), that described such cremation burial pits as "council pits", there are the accounts by Coffin (1963). At his Milford excavations, Coffin found burned and unburned artifacts, greasy dark ash, and burned nuts in large elliptical pits.

Other early assemblages found in burial and habitation contexts are noted by MacCurdy (1962). At the Pine Rock site in New Haven there were several tools of probable Terminal Archaic affinity. Wilcox (1967) more recently assessed the surface material from the Lewis-Walpole site in Farmington as having significant "transitional" materials, noting steatite and broad spears as well as Mansion Inn type cache blades.

By tracing back a collection donated to Yale in 1916, I discovered that in 1903, similar artifacts of probable burial context were discovered in Old Lyme's Salisbury Place. In 1982, I inspected the collection housed in the storage area at the Yale Peabody Museum. Included were projectiles similar to those found at the Griffin site (Pfeiffer 1980 a&b). These exhibited a more local derivation of lithic materials. Quartzite, the primary material to make projectile points, was in much more evidence at Salisbury Place than at the Griffin site. Projectile points were only one category of tool. There were also axes, adzes, two large steatite pot fragments, a cylindrical hammerstone, and a pestle. After reviewing the collection, I suggested that this assemblage may be more recent than the Griffin site. Two possible indications of this were the use of more local material (Pfeiffer 1984) and the higher quality of workmanship seen in the steatite bowl carving technique (Ritchie 1969a:162).

Steatite quarrying activities have been described by Fowler (1951) for Ragged Mountain in northern central Connecticut. Here partial vessels, in various stages of production, and tools for carving such steatite containers were discovered.

Figure 5.3.1 Connecticut Broad Spear Complex Dates and Calibrations
 

Site	Laboratory Lab # Date	Cal BP	Cal BC	Max/Min BP	Max/Min BC
105-34          Beta7808	3610+70	3959,3922	2010.1973	4073-3837	2124-1888
Brodeur Pt     SI-5309	2970+85	3204,3189,3170	1255,1240,1221	3328-2997	1379-1048
105-33         Beta7810	2700+60	2787	838	2859-2758	910-809
Cooper's     Beta5314	2740+90	2850	901	2947-2762	998-813
Schwartz   GX-10850	3335+90	3583	1634	3691-3469	1742-1520
Toelle Rd      SI 5313	3055+120	3328,3292,3268	1379,1343,1319	3389-3083	1440-1134